Unfortunately, Cecil seems to have looked at a number of the same articles that most of the press do, without researching the commentary that occurred AFTER the articles were released.
The Redelmeier & Tibisharani article from the New England Journal of Medicine was beaten to death in the months after its release for an incredible number of methodological mistakes—so many that the authors themselves basically have retracted the majority of their conclusions in subsequent articles. Problems with their study included the fact that they forgot the fact that while cell phone records do accurately reflect the time of the call, there is no such confirmation for the times of the accidents themselves. Based on their results, a cell phone call within five minutes of the time of an accident was counted as being a cell phone related accident. Don’t know about anyone else, but if I personally have an automobile accident, I AM going to be on the cell phone within 5 minutes talking to either the cops, my insurance company, or my wife. These calls have absolutely nothing to do with anyone’s propensity for getting into an accident in the first place. Further, since the average cell phone call lasts for approximately 2 minutes, their results indicated that the cell phone “distraction” effect apparently lasted for 3 minutes AFTER the conclusion of the call, which is unplausible to say the least.
The University of Utah study cited basically states that your average reduction in reaction time is on the order of .08 seconds, or roughly the distance that a car at highway speeds takes to travel a whopping 7 feet. Since another result of being on the phone is a decrease in speed and an increase in following distance (both cited in the same article), the net affect in the event of an accident is an increase in collision speed of 2 feet per second (about 1/3 the traveling speed of the averge walker). I do automotive accident reconstruction, and we’ve modeled it based on the data in the article.
Further, most cell phone research studies stack the deck rather drastically. They put the subjects into an experimental paradigm where they are aware that their performance is being monitored and ask them to drive with their full attention on the roadway. If this condition occurs in normal driving for anyone who has had their license for more than a year or so, it would frankly amaze me—most of us are attending to the driving task while thinking about something other than JUST the driving itself. The cell phone condition, on the other hand, involves an individual multitasking by definition. There is almost always SOME reduction in performance when performing two tasks when compared to one. The “deck stacking” in these studies lies in the underlying assumption that a driver NOT on a cell phone is devoting their ENTIRE attention to the driving task, something not supported by any studies of driver behavior.
When the studies involve a driver having a conversation with a passenger versus a driver having a cell phone conversation, the the results are not usually statistically significant. At that, what an experimenter means by “significant” is not what the general public means by the same term. “Significant” in terms of experimental results means “reliable”, not “substantial”. If research showed that citizens of Los Angeles averaged 1/16 of an inch taller than those of New York, it might be “significant” (i.e., “reliable”), but would hardly be substantial in terms of anything that might make a difference in the way clothing was made.
It is true that cell phone users scan a smaller area to the sides when driving than those not on cell phones, but the same is true for any other task involving substantial mental resources while driving, such as dealing with passengers, eating, drinking, or anything else. The practical effects are just about nil. For some good studies on the same topic, go onto Google and check out the research of Recarte and Nunes, or the study done by the University of North Carolina for AAA a couple of years ago. The Harvard Center for Public Policy also has a good one from the same time frame. Heck, you may be able to dig up some of the ones that I’ve written for IEEE, Professional Safety, or a couple of other technical journals.
In short, the answer is that cell phone use is a distraction while driving, but the jury is still out whether the amount of distraction is significantly different from that of many other “acceptable” activities. On most lists, it ranks about #10, with stuff outside the car being drastically the leading cause (signs, pedestrians, etc.)