Dealing With Uncertainty

In Measurements

• If the uncertainty of a given measuring instrument is specified then that should be used for the uncertainty of the measurement.
• If a measurement is given digitally, for an instrument that has no specified uncertainty, or if a measurement is given with no uncertainty specified, then use one half of the next significant digit as the uncertainty. For example if a digital readout gave 2.63 g and the uncertainty of the instrument was not specified then use 2.63 ± 0.005 g
• If the uncertainty of an analog scale is not known, record the data to one more decimal place than the smallest increment on the scale of the instrument and claim an uncertainty of ± one half of the smallest increment on the scale. (e.g. If we use a scale with 1 mm increments, then we would record the data to the best estimate one more decimal place than the scale, e.g. 37.3 mm. The uncertainty would be one half of the smallest division. e.g. 37.3 ± 0.5 mm.)

In Calculations

• see http://www.saburchill.com/physics/chapters/0070.html
• you will also need to propagate uncertainty for trig functions

In Graphs

• Error bars should be used in graphs, see Graph instruction

Significant Digits

• see http://www.physics.uoguelph.ca/tutorials/sig_fig/SIG_dig.htm

Links

• uncertainty links

IB09 Physics Home