Transcript: NICHD Scientists Identify Risk Factors for Teen Driving Accidents

Bruce Simons-Morton
News Briefing
“Crash and Risky Driving Involvement Among Novice Adolescent Drivers and Their Parents”

Operator: Good morning, ladies and gentlemen, and thank you for waiting.  Welcome to the Teen Driving News Conference.  All lines have been placed on listen-only mode and the floor will be open for your questions and comments following the presentation.  Without further ado, it is my pleasure to turn the floor over to your host, Mr. Bob Bock.  Mr. Bob Bock, the floor is yours.

Robert Bock: Thank you.  Welcome to the National Institutes of Health.  I am Bob Bock, press officer for the NIH's Eunice Kennedy Shriver National Institute of Child Health and Human Development.  One of our scientists, a prevention researcher, Dr. Bruce G. Simons-Morton is here to explain the results of his study, "Crash and Risky Driving Involvement Among Novice Adolescent Drivers and Their Parents."  After describing the study, Dr. Simons-Morton will take your questions.  Dr. Simons-Morton, please begin.

Bruce Simons-Morton: Thanks, Bob.  So, let me first explain to you why the study is unique.  This is the first naturalistic or objective assessment of teenage risky driving.  We used very sensitive data recording devices which we installed in the vehicles of 42 -- the personal vehicles of 42 newly licensed teenage drivers, which were driven by both the teens and their parents so we could compare their behavior.  And the instrumentation included cameras, so we would know what is happening inside and outside the vehicle; GPS, so we know where they are going and we could assess speed; and accelerometers, which allowed us to measure gravitational force events, which is changes in acceleration, such as rapid starts, rapid hard stops, sharp turns and swerves.  And this is our measure of risky driving, and we know from other research that the higher rates of elevated g-force events the more likely a crash in the near future.

And so what we found was that crashes and near crashes -- and near crashes are events just like crashes only an actual contact is averted at the last moment -- among teenage drivers were extremely high right at the beginning of licensure, but they declined rapidly over the 18 months of the study period.  Meanwhile, parent crash rates were much lower.  So, on average teens, over the 18-month period, teens crashed at about four times higher rate and much higher in the first quarter and in the last quarter than adults.  But they did improve over time. 

They did not improve in risky driving.  Risky driving was high relative to parents throughout the entire study.  In fact, it was about five times higher rate among teens than among parents.  Parent drivers almost never had an elevated g-force event and only typically when they would enter the dilemma zone and have to break sharply to make a red light.  But teens do this relatively routinely.

So, what we found was that although with some variability, teens engaged in risky driving sort of style.  They actually got better at driving in this risky manner because their crash rates declined, but they maintained on average a very high level of risky driving.  It's not clear if this behavior is due to intent, that is, they may have driven this way because it's sort of fun, or because they're simply clueless about the risks of such driving.

In related research, we have found that with this population, surprisingly, that after they have a crash or a near crash, they continued driving the same way; they didn't change their driving behavior, which you sort of expect that they would. 

So, I don't know how to explain this except that possibly it's sort of risk compensation.  We know with other complicated still behaviors, like, say, skiing or golfing, or something like that, as you get better you tend to take on a little greater challenge, and this may be exactly what teenagers are doing.  As they gain confidence in their management of the vehicle, they actually maintain sort of a high level of g-force because they're more comfortable rather than learning from it and reducing it.

So, this actually confirms findings from other studies that are self-report or observational, but this is the first objective assessment of teenage risky driving.

You might ask, then, well, what are the implications for prevention?  And sadly it points out the teenage driving dilemma, which is that newly licensed drivers of all ages, but particularly teenagers, are at high risk for crashes early on because they are inexperienced.  It takes a lot of practice to become better, but all drivers eventually become relatively good.  The dilemma is that teens only learn by driving, but the more they drive the greater their risk. 

So, this is the dilemma, and the solution to the dilemma from the policy and prevention point of view is to limit the conditions under which newly licensed teenagers are allowed to drive, either through graduated driver licensing, which is now in place in every state in the nation, but which varies from state-to-state in terms of its strictness, and by parents who can set limits on certain conditions, such as the number of passengers, the use of electronic devices, late night driving, high speed roads.  This may not limit their risky driving, but it limits the conditions that are most ripe for serious crashes, and then over time teens would learn.  And we know from previous research that teens whose parents set limits tend to engage in less risky driving, so we encourage parents to set these kind of limits.

Thank you, and now I'll take questions.

Operator: Certainly.  The floor is now open for questions.  If you do have a question, please press 7 on your telephone keypad.  Questions will be taken in the order they are received.  If at any point your question has been answered, you may press 7 to disable your request.  If you are using a speaker phone, we ask that while posing your question you pick up your handset to provide favorable sound quality.  Please hold while we wait for the first question.  Our first question is from Robert McPherson of AFP.  Robert, please state your question.

Robert McPherson: Yeah, hi, it's Robert McPherson from AFP in Washington, DC, home of the worst drivers in the United States of America.  I had a couple of questions, first of all, about how representative your sample was?  I notice it was concentrated in a couple of towns in Virginia.  One was home of Virginia Tech, so I would imagine there's a lot of teenagers running around there without parental supervision, and the other town sounds like it has a rather higher level of poverty, and also a rather high level of crime relative to other places in the country.  So, how representative do you think these two places would be in the entire country?

And, also, could you elaborate some more your definition of risky driving?  Maybe you did this at the beginning, but I missed it, but are you lumping, say, texting and driving at the same time in the same -- together with, say, excessive speeding on a secondary road?

Bruce Simons-Morton: Sure.  First to the question of representativeness.  It's a little unclear.  These were volunteers and they were from Blacksburg and Roanoke, Virginia, near Virginia Tech, and this was a collaboration with the Virginia Tech Transportation Institute, where we recruited the teens.  So, we don't know for sure, but we do see the Blacksburg/Roanoke area as typical of sort of small cities that are common in the US.  But clearly additional research with other populations is needed to generalize these results.

As to our measure of risky driving, we have a very -- basically, our measure of risky driving is measured by changes in acceleration or gravitational forces exerted on the vehicle.  So, these are basically things like rapid starts or hard stops usually caused by braking late, sharp turns and what we call -- and swerves, or -- and basically the data showed that adults have very, very few of these and most of them are hard stops to avoid running a red light; whereas, teens had a wide variety of these. 

The comparison of teens and parents was for sharp turns, for example, was teens had on the order of 25 to 30 times higher rate than parents on sharp turns.  There is no really good reason to take a turn sharply unless you either are not very good at managing the vehicle or you really enjoy the feeling of going around a turn fast.

The reason that these are associated with crashes is because rapid changes in acceleration reduce the amount of time that the driver has to respond or react to other road conditions, for example, another vehicle entering the road, and they increase the likelihood of loss of control.  So, typically out of a sharp turn a driver may enter into a swerve or a yah, where they overcorrect, and this is a great way to lose control of a vehicle.  So, this is why we think this is a very good measure of risky driving.

Robert McPherson: Thank you.

Bruce Simons-Morton: You're welcome.  Are there any more questions?

Operator: There are no more questions at this time.

Bruce Simons-Morton: I would say just one other thing about risky driving.  In follow-up analyses in a paper that is in press, we look at the relationship between these rates of risky driving or acceleration and the likelihood of a crash or near crash in the next month and found very high association.  Those teens who drove had very low rates of elevated g-force events, also were very unlikely to crash in the next month; whereas, those teens with high rates, and there was variability in this, were much more likely to have a crash in the next month.  So, we think this is an excellent and highly valid measure of risky driving.

Operator: Again, if you do have a question, press 7 on your telephone keypad, please.  All right, it looks like we have another question from Robert McPherson.  Robert, please state your question.

Robert McPherson: Yeah, I'll go ahead.  When I came in, I thought I was the first and only participant so far, so I'll keep you busy for a few more minutes, if you don't mind.

Bruce Simons-Morton: Not at all.

Robert McPherson: The -- let me see, I lost my train of thought here.  I know the insurance industry in its advertising has been saying that -- advising the teens are more likely to have more accidents because simply their brain development is not yet fully completed, which won't be expected until their mid-twenties.  Have you -- I mean, what are your thoughts of that as a contributing factor?  I know that wasn't really part of a statistical analysis, but just anecdotally how do you feel about that?

Bruce Simons-Morton: So, one of the best pieces of information about the effect of young age on crash rates comes from a European study, where they were able to compare the crash rates among drivers who were licensed at 18, and then drivers who were licensed at 21, and then drivers who were licensed at 23 or older.  And what they found was that drivers of all ages, when they first start, crash rates are very high and they decline sharply.  But the youngest drivers had the highest rates and their rate of decline in crashes over time was slower, so that's sort of an age effect.  So, clearly in the US, where teens can get licensed in most states at 16, clearly that age effect may be even more pronounced than it is in Europe.

So, what do we know about 15-year-olds?  Well, if you've raised a teenager yourself.

Robert McPherson: I have not.

Bruce Simons-Morton: Every parent who has raised a teenager knows that at 17 they are more mature than they were at 16.  You don't really need to image the brain to know that maturation is occurring.  We don't have enough information about brain maturation to know what it is about the increased development between the age of 16 and, say, early twenties that might contribute, but it makes sense because we can observe this in teenagers.  They get more mature, they obviously get more mature from age and from experience.

So, one of the great recommendations for reducing crash risk among teens is to delay licensure.  And in states where graduated driver licensing has had the effect of delaying licensure mainly by increasing the amount of supervised practice driving the teens must report before they can get a license, we see reductions in crash rates due to lower exposure.  But it also may be due to some increased maturation.

There is some talk probably mainly idle, about raising the driving age to something as high as 17 or 18.  Currently there is only one state, New Jersey, that has a 17-year-old licensing age, and one study reported that 16-year-olds in New Jersey had much, much lower rates of crashes because they mostly weren't driving or riding with teenagers, compared to the 15-year-olds in the adjacent state of New York, where they can get licensed at 16.

Now, when teens became licensed at 17 in New Jersey, their rates became a little bit higher but not so much that it offset the great advantage of delaying licensure.

So, it's not really possible to image brains while they're driving.  We don't have that technology yet.  That would be really beautiful to see what is going on when they engage in these elevated g-force event rates.  But what you can do is you can do brain imaging and then evaluate aspects of frontal lobe maturation, for example, and then you can evaluate the future driving behavior of teenagers, and I think that kind of research really needs to be done.

Robert McPherson:  Thank you.

Bruce Simons-Morton: You're welcome.  Are there any more questions?

Operator: And there are no more questions at this time.  Again, if you do have a question, please press 7 on your telephone keypad. 

Robert Bock: All right.  If there are no further questions, we will conclude our briefing for today.  If you require a copy of the paper or need anything else, you have my number and contact information from the advisory.  Please get in touch with me.  Thank you.

Operator: Thank you.  This does conclude today's teleconference.  We thank you for your participation.  You may disconnect your line at this time.

top of pageBACK TO TOP