I agree, from a study perspective it would be important to know how they conducted the study and what criteria they used for their data.

"Violations per 100,000 miles driven, expressed as percentage of average"

Where was the study conducted, is it rural, city, combined, time of year, age of drivers, who were the 1.7million drivers? etc., etc., etc. All this plays into how the quality of the data , how it was gathered and the associated results. Also understanding who paid for the study and what they wanted with the results is pretty telling.

A better analysis would be for them to provide a copy of the study and the statistics behind their findings.