Precision How close the measured values are to each other.
Reliability, accuracy, precision and fairness In all types of investigation it is important to consider accuracy, precision, reliability and fairness. Accuracy: When a measurement is close to the true value. Precision: How close the measurements are to each other.
The following example of a target board illustrates the difference between accuracy (getting a bullseye, i.e. the true value) and precision (repeated shots are close to each other).
High Accuracy High Precision
Low Accuracy High Precision
High Accuracy Low Precision
Fig. 2 Comparing accuracy and precision.
Reliability Results that are similar every time the experiment is repeated.
Reliability: The results of an experiment can be called reliable if they are similar every time the experiment is repeated. Fairness: A fair experiment should aim to have only one cause variable that changes. As far as possible all other variables should be kept constant.
Example To show reliability, accuracy, precision and fairness
Two students, Robert and Sarah, were asked to measure the length of a line. Robert used a ruler with centimetres and Sarah used a ruler with centimetres and millimetres. They each made three separate measurements.
This table shows their results: Robert Sarah
19 cm 19.8 cm 18 cm 19.9 cm 18 cm 20.1 cm
The actual length of the line was 20 cm. Reliability: Both students repeated their measurements and obtained similar results so their results are reliable. Accuracy: Sarah’s results were close to the true length of the line so her results are accurate. Robert’s results were not as close to the true length of the line so his results are not as accurate. Precision: Sarah got results that were closer to each other so her results are more precise than Robert’s. Fairness: Two people conducted this experiment and therefore this is another variable. For the experiment to be fair only one person should take measurements.
