Official Car Talk Press Release: August 1, 2001
Tom and Ray Magliozzi, cohosts of the radio show "Car Talk," have accused the American Automobile Association's Foundation for Traffic Safety of issuing a misleading press release on the effects of cell phone use and driving safety.
A press release that garnered extensive national media attention for its surprising results is based on flawed research and is a gross misrepresentation of reality, according to the Magliozzis.
The study-commissioned by the AAA Foundation for Traffic Safety-reported that far less than one percent of accidents involve drivers using cell phones.
The study made use of data supplied by the National Highway Traffic Safety Administration (NHTSA). In that data, distraction is noted only if there is physical evidence at the scene, if it's in the police accident report, or if it's volunteered by the driver. "Be honest," says Car Talk cohost Ray Magliozzi, "How many people are going to say, 'Yes, officer, I was ordering Chinese food on my cell phone when I crashed into that family in the Taurus?' It's just not going to happen."
Even the organization that conducted the study for AAA, along with the National Highway Traffic Safety Administration, says that distracted driving was underreported in the data.
The study's authors, the University of North Carolina's Highway Safety Research Center, note that the database they used "underestimates the role of driver inattention and distraction in crashes." The study authors went on to note that "estimating the true percentage of crashes attributable to various distracting events was not the goal."
Despite this disclaimer by the authors themselves, the press release issued by the AAA Foundation did not mention the authors' disclaimer and simply reported that all distractions contribute to only 12.5 percent of all accidents and that cell phone use accounts for only 1.5 percent of that 12.5 percent-a minuscule affect.
The Magliozzis contacted AAA, suggesting that the release was misleading and was providing ammunition for the cell phone industry's attempts to thwart legislation banning the use of cell phones while driving. "We asked them to issue a new press release that made it crystal clear that the numbers cited in the original release were completely bogus. AAA was totally uncooperative," says Tom Magliozzi.
The AAA director of research, Scott Osberg, Ph.D., suggested in his response to the Magliozzis, "Stick to your sit-down comedy, or if you really care about traffic safety, think about hiring some smart MIT friends to help you understand complex research findings."
"These research findings are hardly complex," says Tom Magliozzi (who, incidentally, also has a Ph.D. in research). "They concluded that distracted drivers get involved in accidents. Duh. The authors of the study say they were simply attempting to develop a taxonomy of distractions. This hardly requires a database. One need only take a drive on any highway in the country to see people reading the newspaper and drying their hair. Why use a database that clearly does not collect such data and then publish percentages?"
Nonetheless, the Car Talk brothers took Osberg's advice and asked an independent research firm to comment on the findings and the press release. Dr. Kevin Clancy, Chairman and CEO of Copernicus Marketing Consulting and research professor at Boston University, had this to say:
"I find AAA's release of the UNC study on 'The Role of Driver Distraction in Traffic Crashes' deeply troubling. To misrepresent findings based on limited data is, simply put, grossly irresponsible. The problem lies not in how the study was planned or executed, but in how the research results are reported and appear to be used.
"The researchers clearly state in their report that their intent in analyzing the NHTSA data was primarily to develop a list of possible driver distractions and other contextual events that lead to accidents. It was not to determine which types of distractions cause the most accidents. To quote from the report, 'These are research results that will be useful in building a broader understanding of driver distraction. The percentages for the different types of distractions should not be used to guide policy development.'
"What is disturbing is that the AAA press release omits this qualification of how to interpret and use the results. Also, presenting findings as actual percentages of each type of distraction among the general population is irresponsible because percentages imply importance to the layperson reviewing findings. This is an incorrect use of the research results.
"The report and release note serious limitations on the data, including underreporting, understatement, and small sample size. They say, 'Present estimates for known distracting events probably understate their true magnitude.' Clearly, the researchers recognized the opportunity for misinterpreting 'preliminary estimates' of driving distractions. This knowledge should have held them and AAA to a higher standard in releasing results to the general public."