Magic Rat said:
I would like to know the type of GPS they used to come up with this.
Handhelds are great for rough guessing with an error of as much as +/-5% from what I have gathered.
The only true way that I see is to have the speedo certified and then check the odometer in a time-speed-distance test.
Put your handheld GPS on the dash, then it's not "handheld", and the accuracy increases. Just kidding of course ;-)
The only way to measure the accuracy of your handheld GPS receiver is to compare the information it provides against a "known good source".
The 5% figure your are referring to likely is in respect to the geographic position coordinates displayed by the receiver. Doing "position averaging" is one technique that can be used to obtain better results from consumer-grade GPS receivers. It may be the case that speed reported by consumer-grade GPS receivers is more accurate than the typical use to report location when stationary, because the speed is derived from differences between locations over time, which, depending on the algorithm used, might take into account a form of averaging.
I have compared my handheld GPS's reported speed against a police laser radar, and they were virtually identical.
However, if you have a "highway test section" nearby, you could use it. To do so you don't need to 'have the speedometer certified'. Do it with a passenger, and use a stopwatch. At the start of the test section start the timing, and at the end stop the timing. Also, note your odometer readings at the start and end, and try and keep at a constant speed(based on your speedometer). If you have a GPS, you might as well use it during the test as well
Once you have the time it took for the test section, that's all you need in order to calculate the "true speed". Take the distance covered by the test section(in Miles or Kilometers) and divide by the time taken, in hours.