The Effects of Music Tempo on Driver Risk Taking

The Effects of Music Tempo on Driver Risk Taking Ashley Hall 1, B. A. & David L. Wiesenthal, Ph. D York University Abstract The effect of music tempo was investigated to determine its effects on risk taking on roadways and performance in a driving simulator. Eighty participants completed a demographic questionnaire about driving history and then were put into 1 of 4 music conditions: no music (control), slow tempo, medium tempo and high tempo. Participants were then tested using two measures of risky driving behaviour: the Vienna Risk-Taking Test Traffic (WRBTV, a subtest of the Vienna Test System (a computer based driving program that measure s risk taking based on reaction times) and DriveSim, a computerbased driving simulator. As predicted on the WRBTV, significant differences between the control group of no music and some of the tempo conditions were obtained with risktaking increasing as music tempo increased. The DriveSim simulator found significant differences indicating that speeding was lower in the slower tempo group compared to the medium tempo condition. Résumé On a examiné l effet du tempo musical afin de déterminer ses effets sur l'agressivité et la performance des conducteurs dans un simulateur de conduite. Quatre-vingts participants ont complété un questionnaire démographique au sujet de leur dossier de conduite. On a ensuite mis les participants dans quatre conditions musicales : aucune musique (groupe témoin), un tempo lent, un tempo normal et un tempo rapide. On a examiné les participants avec deux tests de comportement de conduite risqué : le Vienna Risk-Taking Test Traffic (WRBTV, un subtest du système de Vienna (un programme de conduite automatisé qui mesure le risque en se fondant sur les temps de réaction)) et DriveSim, un simulateur de conduite automatisé. Tel que prédit avec le WRBTV, on a trouvé de différences significatives entre le groupe témoin et les autres groupes, où le niveau de risque augmentait avec le tempo de la musique. Le simulateur DriveSim a noté de différences significatives en ce qui concerne la vitesse de conduite et l écart de vitesse chez le groupe au tempo lent par rapport au groupe au tempo normal, où la vitesse de conduite était plus vite en situation de tempo normal. 1. MUSIC AND DRIVING BEHAVIOUR Research has indicated that music affects arousal levels [1-3]. Dibben and Williamson [4] found that 62% of the drivers reported that music helped soothe them while driving, and 25% reported that it helped them concentrate. Music has also been found useful in reducing stress [3] which suggests that listening to music while driving may reduce stress. Listening to self-selected music may lead to both a reduction in driver Halifax, Nouvelle Ecosse, 8-11 mai 2011 1

aggression [5] as well as reduced driver stress in congested traffic [6]. In another field study, Cummings, Keopsell, Moffat and Rivara [7] found that listening to the radio helped reduce boredom and improved mood which helped reduce the likelihood of crashes. While there is substantial research demonstrating the effects of music on task performance [8-10], there is considerably less research on music s effects on driving performance. A recent study measured participant s reaction time in a driving simulator found listening to music had no effect on their braking time [11]. Many aspects of music such as, genre, volume, tempo, and driver preference need to be considered when examining at music and vehicle performance. Research has begun to focus on the effect of music tempo on driving. Brodsky [1] exposed participants to one of four conditions: no music, and slow, medium, and high tempo music. Participant s performance in a driving video game was assessed along with their physiological arousal were then measured. Participants were twice as likely to make bad judgements and made more vehicle control errors while listening to high tempo music than participants listening to medium or slow tempo music, or no music. Participants were also more likely to drive faster and have more virtual traffic violations when listening to high tempo music compared to slow tempo music or no music. 1.1 THE PRESENT STUDY This study extended Brodsky s study by examining the effects of music tempo on risky driving, using reliable, standardized measures. Risk-taking and performance in a computerized driving simulator were studied by exposing participants to the various tempos used in Brodsky s [1] research. 1.2 Hypotheses This study investigated whether high tempo, medium tempo, slow tempo or no music (control) increases risk-taking behaviour in traffic situations. Hypothesis 1: Individuals in the high tempo music group were expected to exhibit riskier driving behaviour than those in the no music group and were expected to exhibit the greatest risk-taking compared to the other tempo conditions. Hypothesis 2: Music tempo was expected to have a positive linear relationship with risky driving, with the no music control condition displaying the least amount of risky driving behaviour and high tempo related to the greatest amount of risky driving behaviour. As music tempo increased across the different experimental conditions, risk-taking was expected to increase in a similar manner. Hypothesis 3: Participants with a history of traffic violations were expected to demonstrate riskier driving than individuals without history of traffic violations. Method Halifax, Nouvelle Ecosse, 8-11 mai 2011 2

2.1 Participants The participant sample consisted of 80 undergraduate psychology students attending a large, urban Canadian university. There were 20 participants in each condition. Participants were recruited from the subject pool of students enrolled in Introductory Psychology and received course credit for their participation. There were a total of 49 females and 31 males. The mean age of participants was 21.6, SD = 7.279. 2.12 Stimuli In order to avoid confounding with lyrical associations, the selections were entirely instrumental. No top 40 songs were included so that a music preference or familiarity confound was avoided. The musical stimuli consisted of a medley of genres including pop, rock, jazz, blues, funk and country in order arrive at neutral stimuli (a compilation of all genres). The selections of music are the same used by Brodsky [1] and were classified into slow-tempo (40-70 bpm), medium-tempo (85-110 bpm) or fast-tempo (120-140 bpm). Each group (slow-tempo, medium-tempo and fast-tempo) listened to one of four different selections which were randomized for each participant in each tempo class. Participants listened to the music through a headset connected to the computer which played the music at a preset volume. No simulated engine sounds were present in any condition. 2.13 Procedure Prior to the start of the experiment, participants were asked to provide informed consent and then completed a basic demographic questionnaire assessing age, area of residence, driving history, as well as the participant s musical interests (e.g., genres frequently listened to and favourite music). Participants were randomly assigned to one of the four music groups. Participants then completed the DriveSim and WRBTV tasks while they listened to the 4 musical selections in their assigned condition. DriveSim. Participants were measured on the first dependent measure of risk-taking while driving using a driving simulator (DriveSim 3.0, York Computer Technologies, Kingston, Ontario, Canada). The driving task was performed in a virtual driving environment on the computer program DriveSim 4.00 of the York Driving Simulator (York Computer Technologies, Kingston, Ontario, Canada, http://www.yorkdrivesim.com/). The York Driving Simulator (YDS) is a low cost, graphics based, vehicle simulator designed to run on a personal computer. The software can be easily programmed to allow the researcher to manipulate the scenes depicted on the scene that can be monitored at time intervals. Creating scenes which measure participants ability to break, avoid collisions and test their awareness. The type of road that the participant views can also be altered between one or two lanes, speed signs and zones, turns, weather conditions and vehicle dynamics (acceleration, braking, Halifax, Nouvelle Ecosse, 8-11 mai 2011 3

steering).the computer was connected to two 17-inch monitors. One monitor allowed the experimenter to observe the participant s driving performance in detail, with displays of exact speed (km/h) and lane position (m). The second monitor presented the driving simulation to participants [12]. Driving routes were established through previous research experiments employing the YDS [13]. Primary driving functions were essential for controlling the vehicle, keeping it on course. This includes obeying road signs, obeying speed limits and weather conditions. The dependent measures were expressed as mean values over each 3-minute driving route and were: lane deviation, calculated as the deviation in meters of the right wheels of the vehicle from the right edge of the road; speed deviation, measured as the deviation from the posted speed signs in km/h; number of crashes; and a safe zone violation measure, defined as a percentage of time during the driving scenario that the participant would go outside of an artificial box with boundaries placed 1 meter to either side of the center of their lane, and 10 km/h above and below the posted speed signs. The performance variables were sampled from the program 10 times every second. The lower the scores on the driving measures, the better the participant performed. These dependent measures were chosen because they have been used in previous studies and have been reported as good criterion in measuring simulated driving performance [12]. The participants were given two practice trials. Training on the route took place with the software s recording of the amount of lane deviation, speed deviation, crashes, and percentage of time the vehicle went outside the safe zone. In order to qualify in the study, participants had to meet minimum criteria on the practice trials in order to demonstrate that they understood how to operate the driving simulator. The roadway displayed on the screen consisted of a highway with standard Ontario signage. The speedometer is presented on the bottom of the computer screen. The task consisted of maintaining the speed indicated by the speed limit signs, while obeying all street signs and remaining in the right lane unless it is necessary to change lanes. Each participant was told to obey all traffic signs and signals, obey all rules of the road, stay in the right lane unless it was necessary to move to the left hand lane and avoid crashing. It took participants approximately 10 minutes to complete the driving simulator task. Vienna Risk Taking Test Traffic. Participants were assessed on the second dependent variable of risk-taking while driving using the Vienna Risk Taking Test Traffic [14], a recently validated computer administered measure. The WRBTV, a component of the Vienna Test System, measures an individual s willingness to take risks in traffic situations based on reaction times. Participants are presented with 24 different video recordings of driving scenarios (e.g., passing cars, making left and right turns) from the perspective of the driver. The scenes depict different types of road as well as weather conditions. Prior to seeing each of the 24 roadway scenarios, the situation is described in text on the screen. In the second part, participants view the scenario twice. The first time they view the situation, they are told only to observe the situation. The second Halifax, Nouvelle Ecosse, 8-11 mai 2011 4

time, they decide when they would abandon the maneuver by pushing a key on the computer keyboard. The amount of time from the beginning of the scenario to the participant s decision to abort the maneuver by pushing the button is a measure of risktaking behaviour in traffic by using reaction times. The whole testing procedure lasted approximately 40 minutes. The risk taking measure is a total mean score in seconds for the level of risk taken in all scenarios. The higher the total mean score, the less risk actions taken. The lower the mean total score, the more risk actions taken. The DriveSim and WRBTV measures were presented in counterbalanced order. Once participants completed the WRBTV they were fully debriefed and all questions and concerns were answered. RESULTS Before any analysis was conducted, homogeneity of groups was checked. The mean age of participants was 21.6, SD = 7.28. Participants had a mean of driving 4.56 years, SD = 6.10. In the no-music condition, the mean age of participants was 22.6, SD = 8.55 with an average of 5.13 years of driving experience, SD = 6.052. In the slow tempo group, the mean age of participants was 22.85, SD = 10.25, with an average of 5.45 years of driving experience, SD = 10.01. In the medium tempo group, the mean age of participants was 21.40, SD = 5.48, with an average of 4.35 years of driving experience, SD = 3.43. In the high tempo group, the mean age of participants was 19.55 years, SD = 2.24, with an average of 3.33 years of driving experience, SD = 1.89. No significant differences in age or driving experience were found between the groups. 3.1 Music Condition and WRVTB A One-way ANOVA was used to compare the four music condition on WRVTB scores. No statistical significance was found in terms of which measure was presented first: DriveSim or the WRBTV. The absence of order effects permitted the collapse across the two order conditions. Inspection of the Levene statistic (.89, p =.45) indicated that the assumption of homogeneity of variances was satisfied. Assessment of kurtosis and skewness indicated that the assumption of distribution normality was satisfied for all four conditions. Risk Taking (WRBTV) Time in Seconds 9.5 10 8.5 9 7.5 8 No Music Low Tempo Medium Tempo High Tempo Condition Figure 1. Mean scores on the Vienna Test System (WRBTV) and experimental condition Halifax, Nouvelle Ecosse, 8-11 mai 2011 5

The one-way ANOVA indicated, as expected, for the overall WRBTV score there was a statistically significant difference between the four conditions, F (3, 76) = 3.28, p =.045. A linear contrast examined the pattern of the condition means. The contrast value of 4.84 indicated there was a significant linear trend amongst the means of the four conditions, t (76) = 2.92, p =.01. These two results indicated that there was a significant difference between the mean scores, and that the means had an increasing linear relationship. The differences in means and the increasing linear relationship can be seen above in Figure 1. The linear relationship indicated that as risky driving scores increased as music tempo increased. Tukey HSD post hoc analysis was conducted to examine the relations between the four conditions. Post hoc analysis a significant difference (p =.05) between the means of the no music condition (M = 8.32, SD = 1.72) and medium tempo music (M = 9.72, SD = 1.68). There was also a significant difference between the means of the no music condition (M = 8.32, SD = 1.72) and high tempo music (M = 9.76, SD = 1.84). There were no other significant differences. These scores supported the hypothesis that individuals in groups listening to music performed riskier driving behaviours than those not listening to music. 3.2 Driving Offences and the Vienna Risk-Taking Traffic Test A correlational analysis (r = 0.31, p =.25) failed to find a significant relationship between traffic offences and WRVTB scores (i.e., risky-driving scores). 3.3 Music Condition and DriveSim The four groups (no music, low tempo, medium tempo and high tempo music) were compared on their DriveSim scores using a one-way ANOVA. Before conducting the ANOVA, outcome scores were assessed for variance homogeneity and normality. Outcomes scores for speed deviation, speed mean and safe zone had non-normal variance therefore the scores were ranked to reduce non-normal variance as more conventional methods for reducing variance did not work. 3.4 Lane Deviation One-way ANOVA was used to compare the road deviation scores of the four music conditions. Inspection of the Levene statistic (W= 3.95, p =.01) indicated that the assumption of homogeneity of variances was not satisfied. Assessment of kurtosis and skewness indicated that the assumption of distribution normality was satisfied for all four conditions. The original scores were ranked to reduce the heterogeneity of variances. The Levene statistic score of 1.23 (p =.28) indicated the transformed scores satisfied Halifax, Nouvelle Ecosse, 8-11 mai 2011 6

the assumption of homogeneity of variances. The one-way ANOVA indicated there was no significant difference in road deviation scores amongst the four conditions, F (3, 76) = 1.15, p =.33, η 2 =.044. There was a medium effect size, which indicated that 4% of the variance in road deviation is accounted for by variance in music conditions. Time in Seconds 2.8 2.6 2.4 2.2 2 2.3 2.3 No Music 2.4 Low Medium Tempo Tempo 2.6 High Tempo Figure 2. Lane deviation and experimental condition 3.5 Safe Zone One-way ANOVA was used to compare the safe zone scores of the four music conditions on the safe zone. Inspection of the Levene statistic (W= 2.64, p =.06) indicated that the assumption of homogeneity of variances was satisfied. Assessment of kurtosis and skewness indicated that the assumption of distribution normality was satisfied for all four conditions. The ANOVA indicated there was no significant difference amongst the means, F (3, 76) =.08, p =.97, η 2 =.003. There was a small effect size, which indicated that.03% of the variance in the safe zone is accounted for by music conditions. See Figure 3 for a graphical representation of the means. Safezone 39 38 37 36 35 No Music 36.5 37.7 Low Tempo 36.5 Medium Tempo 38.7 High Tempo Figure 3. Percentage of time out of the safe zone and experimental condition 3.6 Speed One-way ANOVA was used to compare the average speed scores of the four music conditions. Inspection of the Levene statistic (W =.14, p =.93) indicated that the assumption of homogeneity of variances was satisfied. Assessment of kurtosis and Halifax, Nouvelle Ecosse, 8-11 mai 2011 7

skewness indicated that the assumption of distribution normality was satisfied for all four conditions. The ANOVA indicated there was a significant difference amongst the means, F (3, 76) = 3.71, p =.02, η 2 =.128. Tukey HSD post hoc analyses indicated that average speed scores in the low tempo condition (M = 65.03, SD = 7.53), were significantly lower than the average speed scores of the medium tempo condition (M = 73.68, SD = 9.92 p =.01, Cohen s d = 0.98). There were no other significant differences amongst the conditions. There was a large effect size for speed mean indicating a large amount of variance in speed mean is accounted for by variance in music conditions. 75 Mean Speed (Km/hr) 70 65 67.5 65 73.7 70.6 60 No Music Low Tempo Medium Tempo High Tempo Figure 4. Mean speed and experimental conditions 3.7 Number of Crashes One-way ANOVA was used to compare the number of crashes of the four music conditions. Inspection of the Levene statistic (W = 4.01, p =.01) indicated that the assumption of homogeneity of variances was not satisfied. A kurtosis score of 5.51 indicated that the distribution of the number of crashes for the medium tempo condition was non-normal. Therefore, the average speed score were ranked, to reduce the variance, and to normalize the score distributions. The ANOVA on the indicated the differences between the number of crashes in the four conditions failed to reach significance, F (3, 76) = 2.28, p =.09, η 2 =.083. There was a large effect size, which indicated that 8% of the variance in crashes is accounted for by variance in music conditions. 3.8 Offences and DriveSim An ANOVA was conducted to determine if there was a relationship between driving offences and scores on the DriveSim, F (3, 76) = 2.28, p = 0.09 indicating there was a non-significant relationship. 3.9 Summary of the Results and Hypotheses Halifax, Nouvelle Ecosse, 8-11 mai 2011 8

It was hypothesized that individuals in the high tempo music group will exhibit riskier driving behaviour than those in the no music group and are expected to exhibit the greatest risk-taking compared to the other tempo conditions (see Table 1 for a summary of the results). Significance was found between the low and medium tempo groups for speed and speed deviation. Participants in the medium tempo group displayed high speed means and more deviance while driving than those in the low tempo group thus demonstrating riskier driving. The second hypothesis was that music tempo will have a positive linear relationship with risky driving, with the no music control condition displaying the least amount of risky driving behaviour and high tempo related to the greatest amount of risky driving behaviour. No significance was found for this hypothesis, although the results were in the predicted direction. The third hypothesis was that people with a history of traffic violations will demonstrate riskier driving than individuals without history of traffic violations. This led to no significant results indicating that individuals with traffic violations do not display riskier driving. Table 1. Summary of results from the DriveSim Dependent Measure DISCUSSION Brodsky [1] examined the influence of music tempo on drivers risk-taking. Individuals in the Dependent ANOVA Probability Effect Size Significance Measure Level Road Deviation F (3, 76) = 1.15 p =.33 η 2 =.044 No significance Safe Zone F (3, 76) =.08 p =.97 η 2 =.003 No significance Speed Deviation F (3, 76) = 3.52 p =.02 η 2 =.122 Significant between low and medium tempo groups Speed F (3, 76) = 3.71 p =.02 η 2 =.128 Significant between low and medium tempo groups Crashes F (3, 76) = 2.28 p =.09 η 2 =.083 No Significance high tempo music group were more likely to make bad judgments and more vehicle control errors while driving. This study obtained significant differences between risky driving scores on the WRBTV. Individuals who listened to the medium-tempo and high tempo music exhibited more risk-taking while driving than slow tempo or no music. Furthermore, as predicted by Hypotheses 1 and 2, there was a linear relationship between risk taking and music tempo. Jonah [15] found that sensation seeking led to increased speeds while driving found that as music tempo increased, so did risk taking. The results also supported the hypothesis that individuals listening to music are more likely to take risks while driving than those who do not. Past research has demonstrated that music tempo can influence an individual s Halifax, Nouvelle Ecosse, 8-11 mai 2011 9

behaviour [1. 16-18]. Edworthy and Waring [16] found that listening to higher tempo musical pieces led to increased drinking rates and Eroglu and colleagues [19] found that higher tempo musical pieces, led to increased walking pace. As in these studies, depending on the music tempo, participants in this study differed in their driving speed. The present data supports Brodsky s [1] finding that individuals who listened to high tempo music were more likely to drive faster than individuals in the slow tempo or no music condition. The present results indicated that individuals who listened to medium and high tempo music had higher average speeds than those who listened to slow or no music. The DriveSim test supported predictions in that participants in the slow music groups had lower levels of speed deviation compared to the medium tempo music group. Results from the DriveSim were not as conclusive as the Vienna Test System, it was only individuals who listened to low tempo music that showed a decreased speed mean and speed deviation. Numerous studies in the past have found that slow tempo; calming music has beneficial effects to individuals behaviours [9-10]. It has been found that music with slow tempos such as background music can have a calming effect on individuals and help improve their concentration by relieving anxiety [8]. This may explain why significance in this driving study was only found between slow tempo and medium tempo groups. This study found that there was a difference in driver s velocity and speed deviation only with the slow tempo group. Having no music while driving could allow participants to drive at their baseline speeds while listening to slow tempo music may relieve anxiety and calm them resulting in a speed decease below their baseline This could be seen in past research where individuals listening to slow tempo background music had lower levels of stress and fewer crashes in traffic situations [7]. There were no differences across groups among the lane deviation, safe zone and number of crashes when groups were compared. Also, there was no statistical difference between individuals with no or varying amounts of driving offences and their level of risk taking while driving. This absence of statistical differences may be attributed to the very few driving offences reported by participants. Driving offences did not vary greatly among individuals (most had 0, 1 or 2 within the past 5 years) which could account for the lack of difference. This ceiling effect had a very low ceiling! Also, there is a difference between the amount of driving offences an individual has and the amount of times they exhibit risky or illegal behaviour while driving. Many individuals drive aggressively and dangerously, exhibiting a lot of risk taking behaviour but are never apprehended by the police, thus they could have a low level of driving offences even when though they commit many of the driving offences. The driving simulator data failed to support the hypothesis that individuals take more risks while listening to music than those who do not listen to music. Although, a more comprehensive measure of the willingness to take risks, the WRBTV found that individuals who listen to music are more likely to take risks in road traffic situations than those not listening to music and that although not every pair of means attained significance, the results were in the predicted direction, that is, as tempo increased, so did risk-taking. Halifax, Nouvelle Ecosse, 8-11 mai 2011 10

4.1 Conclusions: Limitations and future research Currently, there is scant research examining the effects of music listening on driving performance especially risk-taking. There are many possible reasons within music itself (tempo, intensity, type of music, loudness) that may contribute to the effects of music on risk-taking while driving and vehicle performance. The type of music as well as the intensity (loudness) of the music needs to be taken into account on vehicle performance and control. Further research should be conducted examining individuals and the type of music listened to while performing vehicle performance tasks. Perhaps individuals who enjoy a preferred genre of music find it to be more arousing, which may result in more problems in vehicle performance. It could be that the tempo of the music is not as important as the genre of music and an individual s preference for a certain type of music on risk-taking and vehicle performance. Future research should examine whether listening to self-selected music [5-6]. Future studies should examine real life scenarios involving music and driving. In a study conducted by Stutts and colleagues [20], unobtrusive video cameras were mounted in participants cars to monitor the distractions while driving. These included distractions within and outside of the vehicle. Each of the events was recorded along with the frequency and the duration of each event. Similar research should be conducted with video cameras and music. Video cameras should be set up in participants cars similar to the study by Stutts et al. [20] to monitor the music individuals listen to while driving as well as the environment and events taking place in and around the vehicles. Using these cameras would allow the amount of time spent listening to music while driving, the type of music the individual listens to and the effects of the music on the driver s behaviour. Also, diary studies could be used so that participants self-report the type of music they listen to and situations that occurred while driving and listening to that type of music. This would enhance ecological validity and have the impact of real life consequences on the driver. This study also found that the participant s history of driving violations was not related to statistically significant differences in risk-taking while driving. The number of violations, as well as the type of violations varied greatly which could account for the lack of relationship found between violations and risk-taking behaviour. Another area of future research should be conducted where a lengthier history of traffic violations is completed by individuals included how many times in their lifetime they have been stopped, types of violations received and how often they perform legal as well as illegal violations. This would allow us to look at whether some individuals may just perform traffic road violations without being ticketed. Matching drivers with a history of numerous violations with more cautious, ticket-free drivers and then exposing them to the present experimental tempo conditions could answer this question. Halifax, Nouvelle Ecosse, 8-11 mai 2011 11

REFERENCES [1] BRODSKY, W. (2002) The effects of music tempo on simulated driving performance and vehicular control, Transportation Research Part F, 4: 219 41 [2] DILLMAN CARPENTIER, F.R., & POTTER, R.F. (2007). Effects of music on physiological arousal: Explorations into tempo and genre. Media Psychology, 10(3), 339-363. [3] NORTH, A.C., TARRANT, M. & HARGEAVES, D.J. (2003). The effects of music on helping behaviour: A field study. Environment and Behaviour, 36(2), 266 75. [4] DIBBEN, N., & WILLIAMSON, V.J. (2007). An exploratory survey of in-vehicle music listening. Psychology of Music, 35 (4), 571-589. [5] WIESENTHAL, D.L., HENNESSY, D.A. & TOTTEN, B. (2003). The influence of music on mild driver aggression. Transportation Research Part F: Traffic Psychology and Behaviour, 62(2), 125-134. [6] WIESENTHAL, D.L., HENNESSY, D.A. & TOTTEN, B. (2000). The influence of music on driver stress. Journal of Applied Social Psychology, 30, 1709 19. [7] CUMMINGS, P., KOEPSELL, T. D., MOFFAT, J. M., & RIVARA, E. P. (2001). Drowsiness, counter measures to drowsiness, and the risk of a motor vehicle crash. Injury Prevention, 7, 194 199. [8] COCKERTON, T., MOORE, S., & NORMAN, D. (1997). Cognitive test performance and background music. Perceptual and Motor Skills, 85, 1435-1438. [9] RENTZSCH, M., SEIFERT, F., HORNFISCHER, C., SCHREIBER, A. (2008). Melodic segmentation on different musical genres. Proceedings - 4th International Conference on Automated Solutions for Cross Media Content and Multi-Channel Distribution, Axmedis 2008, art.no. 4688043, 3-9. [10] THOMPSON, W.F., SCHELLENBERG, E.G., & HUSAIN, G. (2001). Arousal, mood, and the Mozart effect. Psychological Science, 12(3), 248-251. [11] BELLINGER, D. B., BUDDE, B. M., MACHIDA, M., RICHARDSON, G. B., & BERG, W. P. (2009). The effect of cellular telephone conversation and music listening on response time in braking. Transportation Research Part F: Psychology and Behaviour, 12(6), 441-451. [12] TELNER, J. A. (2007, October). The Effects of second language cellular telephone conversations on Simulated Driving Performance. Paper presented at the Association of Canadian Ergonomics, Toronto, ON. [13] MOLLER, H. J., LOWE, A., KAYUMOV, L., HOSSAIN, N., & SHAPIRO, C. (2002). Can impaired alertness be detected more sensitively using a computerized driving simulator? 252-3. [14] SCHUHFRIED, G. (2006) Wiener Testsystem [Vienna test system]: Vienna Risk- Taking Test [Computer Software]. Vienna, Austria: Schuhfried. [15] JONAH, B. A. (1997). Sensation seeking and risky driving: a review and synthesis of the literature. Accident Analysis and Prevention, 29(5), 651-665. [16] EDWORTHY, J., & WARING, H. (2006). The effects of music tempo and loudness Halifax, Nouvelle Ecosse, 8-11 mai 2011 12

level on treadmill exercise. Ergonomics, 49(15), 1597-1610. [17] MCELREA, H., & STANDING, L. (1992). Fast music causes fast drinking. Perceptual and Motor Skills, 75(2), 362. [18] MILLIMAN, R. E. (1982). Using background music to affect the behavior of supermarket shoppers. Journal of Marketing, 46, 86 91. [19] EROGLU, S. A., MACHLEIT, K. A., & CHEABAD, J. (2005). The interaction of retail density and music tempo: effects on shopper responses. Psychology & Marketing, 22(7), 577-589. [20] STUTTS, J., FEAGANES, J., REINFURT, D., RODGMAN, E., HAMLETT, C., GISH, K., et al. (2005). Driver s exposure to distractions in their natural driving environment. Accident Analysis and Prevention, 37 (6), 1093-1101. Halifax, Nouvelle Ecosse, 8-11 mai 2011 13