random errors affect precision, causing spread about the mean, you cannot get rid of all random errors
ways to reduce random error:
take multiple repeat measurements and calculate a mean, remove anomalies
use computers or data loggers to reduce human error
use equipment with the highest resolution possible
systematic errors affect accuracy, causing all results to be too high or too low by the same amount, you can remove all systematic errors
ways to reduce systematic errors:
calibrate apparatus using an object of known value
if working with radiation, find the background radiation and remove this from the final result
always read from the bottom of the meniscus at eye level
use controls where possible
precision = values which are consistent about a mean, with slight fluctuation
repeatability = the ability of the same person to repeat the experiment with the same equipment and method and get the same result
reproducibility = the ability of a different person to repeat the experiment with different equipment or method and get the same result
resolution = the smallest change in the quantity being measured that gives a recognisable change in reading
accuracy = measurements close to the true value
uncertainty = upper and lower bounds in which you would expect to find the value
absolute uncertainty is given as a fixed quantity
fractional uncertainty is given as a fraction of the measurement
percentage uncertainty is given as a percentage of the measurement
to reduce percentage and fractional uncertainties, measure larger quantities of the thing
readings are when one value is found, measurements are when the difference between two readings is found
uncertainty in a reading is plus or minus half the smallest division
uncertainty in a measurement is the sum of the uncertainties in the readings which make it up
uncertainty in digital readings is either given or assumed to be plus or minus the smallest possible reading
to reduce uncertainty in a ruler, fix one end of it as this reduces the number of measurements required
uncertainties should be given to the same number of significant figures as the data
when adding or subtracting data, add the absolute uncertainties
when multiplying or dividing data, add the percentage uncertainties
when raising data to a power, multiply the percentage uncertainty by the power
uncertainties can be shown on graphs using error bars
a line of best fit or a line of worst fit needs to go through all the error bars on the graph
to find the uncertainty in a gradient, draw two lines of worst fit, a steep one and a shallow one, calculate the gradients of each, the uncertainty is the difference between these