"I worked it out by walking a distance measured using google maps then dividing that by the number of strides recorded."
That's what I did with my old mechanical one, but the trick is to make the same measurement several times and see whether you get the same answer. The first point with any measuring instrument is that if it's not repeatable it can't be accurate. Mine was giving a different answer every time.
"Crun - love to know more about the accuracy of cals burned - i.e. was it over or under?"
"So if I am walking and using a pedometer for calorie information it is only 35% accurate, is that what you are saying."
If the accuracy of an instrument is specified by the manufacturer as 5% say, then that means that if the true value being measured were 100, then they're guaranteeing that the reading from the instrument will be somewhere between 95 and 105, but the point is that you can't establish the manufacturer's specified accuracy by measuring. If you were to pick up an instrument specified as 5% accuracy, it might be within 2% , but if you were to conclude from your measurement that the accuracy were 2% you'd be making a mistake, because another example of the same instrument may be as much as 5% out an still be within the manufacturer's specification. All you would be able to say from your measurement is that the accuracy is no better that 2%.
So for our pedometers, if the accuracy were 65%, the reading could lie anywhere between 35 and 165. Which say that the figures they are quoting are averages, so the measurements for some pedometers must have been worse than 65%, and as we've seen, another one off the shelf could easily be worse than the measurements.
Roughly translated, it means they're about as much use as a chocolate teapot for measuring calories, but the step counting doesn't look so bad for walking and running.