Accuracy in measurement is based on *relative error* and number of *significant digits.* Relative error is the absolute error divided by the calculated (or estimated) value. For example, if a person expects to spend $10 per week at the local espresso bar, but actually spends $12.50, the absolute error is 12.50 - 10.00 = 2.50; the relative error then becomes (2.50 / 10) = 0.25 (to find out the percent, multiply by 100, or 0.25 × 100 = 25 percent of the original estimate). Significant digits refers to a certain decimal place that determines the amount of rounding off to take place in the measurement; these numbers carry meaning to the figure’s precision. But beware—accuracy in measurement does not mean the actual measurement taken was accurate. It only means that if there are a large number of significant digits, or if the relative error is low, the measurement is more accurate.