DUM: DRIVING UNDER THE MUSIC
A First-Year Study of the Effect of Music on the Braking Reaction Time Among Male and Female Teenage Drivers
6008 SE 56th Court
Oklahoma City, Oklahoma 73135
Mount Saint Mary High School
2801 South Shartel
Oklahoma City, Oklahoma 73109
Appendix A: Informed Consent Form (4B)
Appendix B: Slide Pictures
Appendix C: Individual Subjectís Data Record Form--DOUBLE SIDED
Appendix D: Procedure
Appendix E: Complete Data Table
Appendix F: Average Reaction Time in Seconds By Trial
Appendix G: Average Reaction Time in Seconds For All Trials Combined
Appendix H: Deviation from Control in Seconds By Trial
Appendix I: Deviation from Control in Seconds For All Trials Combined
Appendix J: Correct Responses Percentage For All Trials Combined
Appendix K: Spearman Correlations Data Table
Appendix L: Spearman Correlations: Subjects at Each Statistical Ratio
DUM: Driving Under the Music. A First Year Study of the Effect of Music on the Braking Reaction Time Among Male and Female Teenage Drivers
Hopper, Larry J.
6008 SE 56th Court, Oklahoma City, OK 73135
Mount Saint Mary High School
Teenage drivers have been under extreme scrutiny for many years. Sixteen- and-17-year-olds account for only 2% of all drivers, but they are involved in almost 11% of all crashes (Cohen 80). Many factors cause this, but one overlooked problem is how easily teens become distracted by friends and music. If music proved to be a significant distraction, then listening to music could be initially restricted through graduated licensing until teens gained substantial driving experience.
Thirty-two teenage subjects (19 males, 13 females) were tested with a homemade brake machine which triggered a high-memory stopwatch. Subjects reacted to slide images by either braking or choosing not to while not listening to any music (control condition) and while listening to music played at 70, 85, and 100 decibels. All the music levels had five slides that required a response, and five that did not.
Data revealed that subjects braked fastest at the control level and slowest at 85 decibels. The most braking errors occurred at 100 decibels. Both sexesí results were identical at the control and 70 decibel levels. Males committed more braking errors at 85 decibels while females made more at 100. Femalesí reaction times were much slower than malesí at 85 and 100 decibels, especially at 100 where the difference was almost 0.13 seconds. The Spearman rs Rank-Correlation Test displayed no statistical significance for the subject sample but did for almost half the individual subjects.
The results partially support that volume did slightly affect teenagersí braking reaction times but not significantly. However, females did not brake faster and more appropriately than males as hypothesized.
"Why are teens the worst drivers? Because too many are easily distracted risk takers. All too often, they fail to see a dangerous situation developing as they fiddle with radio dials, get swept away by their favorite songs, or pay more attention to their passengers than to oncoming traffic (Fields 74)." Teenagers listen to music constantly, whether it be at home doing homework or while driving a car. They attempt to devote their attention to both the music and their driving, which could be a dangerous combination. If music, especially when loud, caused significantly slower reaction times, then precautions could be taken to eliminate the problem.
Teenage drivers have been under extreme scrutiny for many years, and statistics reveal a rather large recent rise in teenage motor vehicle deaths, the leading cause of death for 15 to 20 year olds (Steering Committee 1). From 1992 to 1996, teenage fatalities gradually rose from 5,215 deaths a year to 5,805 (Insurance Institute for Highway Safety 4). Those figures may not seem extremely large, but they account for only the fatalities. Statistics show that five out of 20 (25%) teenage drivers will be involved in an auto accident, and over one in 20 (> 5%) will be seriously injured or killed in an accident (Steering Committee 1).
Why is the teenage crash rate so high? It can be partly attributed to teenagersí lack of experience, likeliness to take risks, nighttime driving, drunk driving, and low rate of seat belt usage (Save A Teen, Inc. 1). However, one factor that often gets overlooked is how easily teens are distracted by their surroundings, especially by their friends and music. This causes them to be inattentive while driving, which greatly alters their judgment (Preparing Your Teen to Drive 5). One Internet source listed ways to reduce the chance of an accident. Under it fell the necessity of avoiding distractions while driving, including loud music (Save A Teen, Inc. 2). Therefore, researching the effects of music on braking reaction time among teenagers seemed to be a promising and intriguing subject even though little previous research was found on the topic.
Through this research experiment, it was hoped that the data would conclusively determine whether music slowed reaction time or not. If it did, then it would also be worth testing whether teens could still listen to soft music and not have a slowed reaction time.
Quick reflexes, or reaction times, are extremely important when driving a car (Hjelmeland 15). Reaction time is the sum of the time that is required for the brain to: (1) receive information from the senses, (2) make a decision as to what action to take next, (3) to transmit this decision from the brain to the muscles needed to carry out the reaction, and (4) having the muscles make the response (Scotti 69). The most critical of all these steps is the second because that is where the decision about what reaction to take is made. The standard reaction time for a healthy person is 0.66 seconds (Scotti 71).
Reaction time was determined by the researcher to be most easily tested through braking. Since tailgating, or following too closely, is the leading driving mistake in teenage accidents (Preparing Your Teen to Drive 5), it also seemed practical to use braking reaction time.
The hypothesis of this research paper stated that music would cause a slower reaction time among teenage drivers as opposed to not listening to music. However, music at a low volume should not cause a significantly slow reaction time. Females should brake faster and more appropriately while listening to music than males since their insurance rates are lower .
The purpose of this project was to discover if music was a major distraction to teenagers while driving and if a low volume could be acceptable. Another purpose was to discover if there was a difference in the reaction times and responses between males and females. If music proved to be a significant distraction, then listening to music could be initially restricted through graduated licensing (adopted in eight states), "a three-tiered system of limiting privileges that become more generous as a teen gains experience (Cohen 80)." However, if no consistent effect was noticed, then other areas of commonly overlooked teenage driving errors could be considered for future research.
Building the Brake Apparatus and Other Preparations
Before experimentation could begin, numerous tasks were completed in preparation for testing. First, the brake pedal testing apparatus was built from scratch. It was an original, homemade device that was used to measure each subjectís braking reaction time in hundredths of a second. The machine consisted of a large wooden base measuring 48 cm by 51 cm. An accelerator pedal was attached to a six- inch hinge on the right. It was located 25 cm from the front of the base and 2.5 cm from the right side. The bottom of the pedal was 5 cm above the base (like in most cars) and this "hinge pedal" was attached to a screw and a spring which fit in a wooden block. The brake pedal was attached to an eight-inch hinge on the left. It was located 20 cm from the front, and 18 cm from the right side. The bottom of the pedal was 9 cm above the base (like in most cars), and this "hinge-pedal" was attached to a screw and a spring which fit in a block. However, this brake hinge also had another spring attached on the back near the top, and this spring depressed the button on the stopwatch. The stopwatch was securely housed in a box behind the pedal. Figure 1 below includes a picture of the apparatus while it was being built.
After the brake pedal apparatus was built, the second task was to obtain a slide projector and a compatible slide synchronized tape recorder. The recorder had to have adjustable volume settings and be able to connect to headphones. Then, using slide film, 20 brake reaction slides were taken of different driving situations. Situations used in this project included five cues: a red traffic signal, car braking immediately ahead, stop sign, car backing out of a driveway, and a child running into the street. Then, 20 more slides to which people should not brake were taken with the camera. These "dummy slides" included anything from open roads to a building or a green traffic signal.
Next, 72-second clips for the music tape were chosen from the beginning of five different well-known recent songs. The songs chosen for this project were the top five songs on the Top 40 chart for the week of December 6th, 1998, the week that the actual experimentation began. In this order, the songs on the tape included: "One Week" by the Barenaked Ladies, "Are U That Somebody" by Allyiah, "Save Tonight" by Eagle Eye Cherry, "Thank U" by Alanis Morisette, and "Jumper" by Third Eye Blind.
The magnetic time pulses were then set at random on the music tape using the synchronized tape recorder. These pulses advanced the projector the same way for each subject and were placed in a range from seven to twelve seconds apart. Eight pulses were needed for each song clip, or trial. Once the pulses were set, it was necessary to determine when the 40 actually occurred using the high-memory stopwatch. The 40 pulses were timed five times using the stopwatch and the average of them was calculated to figure these "master times" out. Dr. Blas Espinoza-Varas from John W. Keys Speech and Hearing Center was consulted to set the three safe sound levels on the volume control dial of the slide synchronized tape recorder at 70, 85, and 100 decibels. Dr. Espinoza was kind enough to lend his time, his soundproof room, decibel meter and advice for this project. He stated that all sound levels used "were completely safe at the time they would be sustained for (Espinoza Interview)." Finally, the true experimentation could begin.
Thirty-two (19 males, 13 females) teenage test subjects ranging in age from 15.24 years old to 19.15 years old were recruited. The mean age of the subjects was 16.23 years old. All subjects had a basic knowledge of driving and signed informed consent forms (4B); parents signed for minors under 18. This form is labeled
Appendix A. The subjects were not divided into groups as they were tested in the same manner, but there were control and experimental conditions for each subject. The control condition was to have no music at all, just the normal background noise of 40 to 45 decibels. The experimental conditions were the three separate decibel levels. According to a psychologist, having just one group of subjects was valid since there were experimental and control conditions. Also, at least 20 subjects would be needed to generate statistically valid results as each subject provided five response times for each of the music levels, including the control (Gronlund Interview).
One subject was tested at a time. Upon being seated at the testing apparatus, each subject was shown pictures of the five brake reaction stimuli scenes and informed to only brake to slides of those pictures. These pictures may be viewed in Appendix B. Each subject was also informed that there were 20 brake reaction slides and 20 dummy slides to which they should not brake. The subject was then instructed to place the headphones comfortably over both his/her ears, and they were told how to use the brake and acceleration pedals. Once the subject said that they were ready, the stopwatch and the tape were started simultaneously. Then the stopwatch was placed in its box and checked to make sure that the spring would activate it. As mentioned previously, the synchronized tape recorder ensured that all subjects were tested the same. A one-second dissolve rate was used on the slides. The stopwatch had to have a high memory capable of storing at least 20 brake reaction times.
While the music tape played in the headphones, the researcher altered the volume levels in the same order for each subject. The researcher noted whenever the subject depressed the brake pedal when they should not have and noted when the subject forgot to brake when they should have. The test itself consisted of 40 slides and five trials (one trial per song) and lasted six minutes (not counting the 30 second delay at the beginning). Each subjectís responses were logged onto their individual record form. The data form is located in Appendix C.
Once the six minute test was completed, the subject was asked questions about how the test went. Then, the subjectís reaction times were retrieved from the stopwatchís memory and logged onto the subjectís data sheet. His/her actual reaction times were the master times subtracted from these response times. These were all entered into a spreadsheet.
Once all the subjects were tested, the data was organized and analyzed. There is a step-by-step procedure located in Appendix D.
When the data was being analyzed, it became apparent that subjects had an extremely slow reaction time on the very first reaction. This reaction was to the control condition, and it was more than 0.2 seconds slower than the next slowest reaction. Through consultation with teachers and Dr. Espinoza-Varas, it was decided that this reaction should be excluded since the reason for its abnormality was due to the subjects not having practice. Even though the other reaction times in Trial 1 were perfectly normal, they were also discarded because there was not a control condition to which they could be compared. Thus, no results in Trial 1 were included, but the results for Trials 2 through 5 have been analyzed and were more than sufficient. A full data table in Appendix E displays each braking reaction time, correct responses, and statistics (Spearman rs Rank-Correlation Test) for each individual subject.
Results showed that on average, subjects had the fastest braking reaction time at the control level when no music was played. On average, they took 1.5508 seconds to react at this level. Brake time was second fastest at the highest level, 100 decibels, at an average time of 1.5657 seconds. The next fastest level was the first music level, 70 decibels, where subjects had an average reaction time of 1.5792 seconds. Finally, subjects had the slowest reaction time at the second level, 85 decibels. It took subjects 1.5817 seconds to brake at this level. In other words, the data did vary between each of all the four valid trials, but when they were averaged, the results showed the slight difference between each of the levels with the control level producing the fastest reaction time and the 85 decibel level producing the slowest reaction time.
Appendices F and G are line graphs that show correlations for each separate trial and for all the trials combined.
Males and females had very similar reaction times at the control and 70 decibel levels. Females were less than 0.01 seconds slower at both as can be seen in the table below. However, at the 85 decibel level, males and females performed in opposite directions. Males actually improved their times at 85 decibels and were over 0.05 seconds faster than the females, whose times regressed. This trend continued at the highest level (100 decibels) where the males had their fastest braking reaction time (1.5147) and the females had their slowest (1.6402). This difference of almost 0.13 seconds is quite significant and should not be taken lightly since at 60 miles an hour a car moves almost 100 feet per second. See the table below for the exact numerical figures.
|REACTION TIMES IN SEC. FOR TRIALS 2-5 COMBINED|
|70 dB||85 dB||100 dB|
Overall, the results did not deviate much from the control times, but each level had a slower time than the control. Within each trial, however, the results had fairly large positive and negative deviations. See Appendices H and I for graphs that show these deviations for each trial and for all the trials combined.
A percentage for appropriate braking responses was also obtained. Anytime a subject failed to brake when they should have or used the brake when they should not have, the percentage was lowered. There were eight possible responses for each subject; 256 possible for all the subjects combined at each sound volume. There were 251 out of 256 (98.05%) correct responses at both the control and 70 decibel levels. At 85 decibels, the percentage rose to 98.44% (252 out of 256), but at 100 decibels, the percentage dipped to 94.92% (243 out of 256). There were differences between males and females at every level except the control level. At 70 decibels, males had a slightly higher percentage, but at 85 decibels, females were less likely to make an error as they actually were 100% at that level. However, females had extreme difficulty at 100 decibels, having 97 out of 104 correct responses for 93.27% while males had 146 out of 152 correct for 96.05%. Appendix J shows the graph displaying this information, and the table below indicates all the percentages.
|CORRECT RESPONSE % FOR TRIALS 2-5 COMBINED|
|70 dB||85 dB||100 dB|
The Spearman Rank-Order Coefficient of Correlation (Spearman rs) was used as the statistical significance test for this project (Wright 252). The scale on this test ranged from -1.0 to 1.0, with -1.0 representing a strong negative correlation, 0 representing no correlation, and 1.0 representing a strong positive correlation. The formula was as follows: rs = 1- 6S D2/ n(n2- 1). The test revealed that there was a correlation of -0.06 for all subjects, meaning that there was an extremely small negative correlation. Thus, as volume increased, reaction times improved. The femalesí data showed a correlation of +0.11, a very small positive correlation meaning that femaleís reaction times were slower at higher volumes. The malesí data indicated a correlation of -0.18, a small negative correlation that is almost somewhat significant.
Even though none of these figures were statistically significant, many subjects performed within strong statistical significance. The data table for these calculations is located in Appendix K, and a graph simplifying this is located in Appendix L. The graph shows that 6 of 19 males had extremely strong negative correlation (as opposed to 2 of 13 females) and that 5 of 13 females had extremely positive correlation (as opposed to 2 of 19 males). Many subjects of both sexes also displayed little or zero correlation. These statistics indicate that although the correlations were not statistically significant for the group as a whole, they were statistically significant for many individuals. It also shows how more males have negative correlations than females as the time reaction graphs suggest, and how more females have positive correlations than males.
An error in this experiment that has already been mentioned is how the subjects did not have an opportunity to practice braking. However, this problem was corrected by eliminating the results of Trial 1. Possible errors in this experiment include how the simulation may not have mimicked driving completely and when timing the master times for the slide reaction show (although the slides were timed over 5 times).
One factor in this experiment that was not controllable but may have altered the results is how some subjects were more "motivated" to do well than others. Another is that some people may not have liked a certain type of music, thus causing them to brake more slowly on that particular song.
If this experiment was repeated, the researcher would definitely include a built-in practice trial because of the reasons already stated in this section of the paper. The researcher would also have tested fewer subjects to reduce stress and would have tried to balance the number of males and females for the sake of the results.
Further investigations that could be conducted include: "Does the use of cell phones while driving cause drivers to have significantly slowed reaction times and make more incorrect driving decisions?", and "Do various types of music cause people to drive differently?"
The results of this experiment appear to partially support the first part of the hypothesis. Music volume did slightly affect teenagersí braking reaction times but not significantly. However, the second part was not correct as females did not brake faster and more appropriately than males.
Subjects tended to brake the fastest when no music was played (control condition: Level 0) and tended to brake the slowest when music was played at 85 decibels (Level 2). The control, 70 decibel and 85 decibel levels all had similar appropriate response percentages, but they dropped off significantly at the 100 decibel level.
Males and females had identical braking reaction times and correct response percentages at the control and 70 decibel levels. Males applied brakes faster at the 85 decibel level but had a worse correct response percentage than females. However, the greatest differences occurred at the 100 decibel level. Males exhibited their fastest braking reaction times at that level while females had their slowest. Females also struggled to brake appropriately at 100 decibels.
The first acknowledgment is to my parents who assisted me tremendously throughout the project. Their transportation, assistance, financial aid, and moral support were essential to the success and advancement of this project.
A second acknowledgment is to Professor Scott D. Gronlund, the Associate Professor of Psychology at the University of Oklahoma and Dr. Blas Espinoza-Varas of the Keys Speech and Hearing Center. Dr. Espinoza-Varas was kind enough to lend his time, his soundproof room, decibel meter, and advice for this project.
The 32 test subjects are well deserving of my third acknowledgment as each one was willing to volunteer ten to fifteen minutes of his/her time for this research.
My fourth acknowledgment is to Delona Elsberry and Diana Mack from the Media Center of Moore Public Schools who allowed me to borrow a slide projector.
My final acknowledgment is to the Mount Saint Mary faculty, who have been very supportive and helpful throughout this long process.
1. Cohen, Warren, "How to Reduce Teensí Road Accidents: ĎGraduatedí Licenses," U.S. News and World Report, Dec. 29, 1997/Jan. 5, 1998, p. 80.
2. Espinoza-Varas, Blas (Dr.), interviewed by Larry Hopper Audiology. John W. Keys Speech and Hearing Center, 825 NE 14th St., Oklahoma City, Oklahoma. 3:15 P.M., December 7, 1998 and 4:30 P.M., January 5, 1999.
3. Fields, Barbara, "When Kids Drive...," Better Homes and Gardens, Sep. 1997, p.74.
4. Gronlund, Scott D., interviewed by Larry Hopper Psychology Department. Dale Hall Tower, 455 W. Lindsey Room 739, Norman, Oklahoma. 3:45 P.M., October 14, 1998.
5. Hjelmeland, Andy. The Facts About Drinking and Driving. Crestwood House,
New York, New York; 1990.
6. Insurance Institute for Highway Safety. http://www.hwysafety.org/facts/teens.htm
(28 Sep 1998).
7. Preparing Your Teen to Drive. State Farm Mutual Insurance Company, Bloomington, Illinois; 1990.
8. Save A Teen, Inc. http://www.save-a-teen.com/tips.htm (28 Sep 1998).
9. Scotti, Anthony. Driving Techniques for the Professional and Non-Professional. Photographics Publishing, Ridgefield, New Jersey; 1995.
10. The Steering Committee. http://www.steeringcommittee.com/ (28 Sep 1998).
11. Wright, R.L.D. Understanding Statistics: An Informal Introduction for the Behavioral Sciences. Harcourt Brace Jovanovich, Inc., New York, New York; 1976.
APPENDIX A: Informed Consent Form
NOTE: THIS IS AN I.S.E.F. FORM AND WAS NOT AVAILABLE AS A WORD DOCUMENT, NOR WAS IT SCANNED AND TURNED INTO AN IMPORTABLE GRAPHICS FILE.
APPENDIX B: Slide Pictures
NOTE: THESE PICTURES OF THE SCENES THE SUBJECTS HAD TO RESPOND TO ARES NOT AVAILABLE AS A WORD DOCUMENT, NOR WERE THEY SCANNED AND TURNED INTO AN IMPORTABLE GRAPHICS FILE.
APPENDIX C: Individual Subjects Data Record Form
NOTE: THIS ISA FORM WHICH IS NO LONGER AVAILABLE AS A WORD DOCUMENT, AND IT WAS NOT SCANNED AND TURNED INTO AN IMPORTABLE GRAPHICS FILE.
APPENDIX D: PROCEDURE
I. Building the Apparatus and Calibrating the Test
1. Build a sturdy brake pedal testing apparatus. This machine consists of a large wooden base along with six and eight-inch hinges attached to springs. Each hinge is attached to a brake or acceleration pedal. A box is also attached to hold the stopwatch which will be in front of the brake pedal. The stopwatch is activated whenever the spring from the brake pedal depresses the stopwatch button.
2. Obtain a slide projector and a slide-synchronized tape recorder. Connect the recorder to the slide projector and to headphones. The recorder must have an adjustable volume control.
3. Using slide film, take 20 brake reaction slides of different driving situations. Those used in this project included a red traffic signal, tailgating car, stop sign, car backing out of a driveway, and a child running in the street. Then, take 20 more slides to which people should not brake ("dummy slides"). Scenes included can be anything to which a driver would not brake, such as an open road, a building, or a green traffic signal.
4. Choose 72-second clips for the music tape from five different well-known recent songs. A good way to do this is to choose the top five songs on the Top 40 chart for the week when testing is begun.
5. Set the time synchronization pulses randomly on the music tape using the synchronized tape recorder. These pulses will advance the projector the same way for each subject and should be placed anywhere from seven to twelve seconds apart. Eight pulses are needed for each song clip or trial. Once the pulses are set, determine when the 40 pulses actually occur using a high- memory stopwatch. Measure the 40 pulses five times, and calculate the average of these times to figure the "master times" out.
6. Consult with an audiologist and set three safe sound levels on the volume control dial of the slide-synchronized tape recorder. The levels used in this experiment were 70 decibels, 85 decibels, and 100 decibels.
II. Testing Subjects
7. Assemble a sample of 20-40 teenage subjects who are at least 15 but are no older than 19 and have a basic knowledge of driving. Have each subject and one of their parents (if the subject is under 18) sign consent form 4B.
8. Test one person at a time. Seat them at the testing apparatus and show the subject pictures of the five brake reaction responses and inform them to only brake to those pictures. Tell them that there are 20 of those slides to which they should brake and that there are 20 slides unlike those to which they should not brake. Then instruct the subject to place the headphones over his or her ears and that he or she is to place their right foot on the acceleration, or right pedal. Inform them that whenever a braking response slide appears ten feet in front of them, they need to press the brake, or left pedal as if they really were driving, but return the foot to the right or acceleration pedal after they have depressed the brake.
9. Start the stopwatch at the same time you start the tape. Then place the stopwatch in the box and check to make sure that the spring activates it.
10. As the music tape plays in the headphones, alter the volume levels in the same order for each subject. Note whenever the subject brakes when they should not brake and when the subject forgets to brake when they should. The test itself consists of 40 slides and five trials. Each trial has eight slides. Two are control conditions; there is no music, and one of these two requires a brake response while the other is a dummy. Six are experimental conditions, two for each of the three music levels (70, 85, and 100 decibels). There is a response and dummy for each.
11. Once the six minute music tape is over, and the test is completed; ask the subject questions about their comfort levels, birthdates, and how the test went. Then, retrieve the subjectís reaction times from the stopwatchís memory and copy them down on the subjectís data sheet. Their actual reaction times are the master times subtracted from these times, and they will be calculated using spreadsheet software later.
12. Repeat steps 8 thru 11 for each subject.
13. Once all subjects have been tested, organize and analyze the data.