Slashdot Mirror
Talking On the Phone While Driving Not So Dangerous After All
Dorianny writes "New research which takes advantage of the increase in cell phone use after 9pm due to the popularity of 'free nights and weekends' plans showed no corresponding increase in crash rates (PDF). Additionally, the researchers analyzed the effects of legislation banning cellphone use, enacted in several states, and similarly found that the legislation had no effect on the crash rate. 'One thought is that drivers may compensate for the distraction of cellphone use by selectively deciding when to make a call or consciously driving more carefully during a call.' Score this a -1 for common sense."
You fuckers need to keep your hands on the God damn wheel.
You have limited infoprocessing resources. You spend some on a conversation, its less for driving. Conversations can be more distracting than ethanol. Its pretty simple. I've told my wife and kid to shut up when I'm concentrating on a new route. Know your limits.
Talking and texting while driving was made illegal. Accident rates didn't change. That doesn't say anything about how dangerous it is to talk or text while driving. Instead, it just says that the law is sporadically enforced, if at all, and universally ignored by drivers. Accident rates didn't change because talking/texting while driving rates also didn't change.
I question how much free minutes changed calling patterns, too. I suspect cell phone companies offered that feature knowing there would be little or no change in calling patterns and they would continue to make nearly all the money they already were before the change, indicating that people aren't taking advantage of free minute time windows.
There's also another way to interpret the data—that the negative effects of using the phone more after 9 P.M. for fully awake drivers are cancelled out by the positive effects of ongoing interaction with another person helping keep sleepy drivers more alert. If this is the case, then banning cell phone use might actually cost lives....
Check out my sci-fi/humor trilogy at PatriotsBooks.
More likely those people are just not representative of drivers using cell phones. You notice them more, because of selection bias.
Most cell phone drivers are the ones sitting in some random lane, not changing lanes, driving slow and making everyone pass them. They are sitting at red lights after the green, and letting people pass when they should go.
"I opened my eyes, and everything went dark again"
Actually a friend of mine tells an amusing story of being in a class in HS where the teacher brought out the alcohol and driving stats and asked the class "What do these stats tell you?"
Apparently the teacher didn't like it when he raised his hand and said something which I actually believe to be true: "It takes about 10 years to learn how to drive a car well".
I would have laughed at you had you said that to me when I was in my early 20s. At this point, I would smack my 20something self for being stupid.
"I opened my eyes, and everything went dark again"
Just to list a few:
For starters this is a retrospective, observational (being generous here) cohort study.
I'd like a bit more technical detail on how they ensured that they were measuring mobile calls from cars (they have assurance from the telecommunications company)
They note a 7% rise in what they believe to be car mobile phone calls at 9pm on Monday to Friday on a background of steadily decreasing phone calls from 8pm to 10pm, and they don't mention whether this spike is statistically significant.
The spike in the rise of mobile car use is of a maximum of 1/2 hour before the level reaches pre-9pm levels, and continues to decrease. This interval is short - to notice an effect the recording of the car accidents in their source would have to be pretty precise. Any errors in the reporting of car accidents is probably going to make a 30 min window period difficult to measure.
They haven't analysed the variation in traffic at different times in the evening, which makes comparison at different time periods difficult. If the traffic is less after 9pm, the rate of accidents per car could be higher.
But the main problem is:
To show 'no effect' you need to ensure that your study is powered to make this observation - which they have not done. A 7% rise in mobile usage over 30 minutes would need ?how many crashes to give a statistically significant result that rises above the noise.
To be fair, they mention some of these issues as caveats, but I'm not sure they had enough statistics input for this paper. I would like to see the confidence intervals, how they were calculated, what software was used and what the p-values are. There should be a statisticians name on the paper. Certainly, you can't conclude that mobile phones are not dangerous while driving - you can only say that they found no evidence to show this in this particular study.