A worked example
Measuring 48 when the accepted value is 50 gives a percent error of 4%.
Frequently asked questions
Why use the absolute value of the difference?
Percent error measures how far off a measurement was, not which direction — taking the absolute value means an experimental result that's too high gives the same percent error as one too low by the same amount.
What counts as a 'good' percent error?
It depends entirely on the field and the measurement — some lab contexts expect well under 1%, while rougher estimates might consider 10-20% acceptable. There's no universal threshold.