Reliability, Validity and Accuracy

Validity – (Trustworthy)

An experiment is valid if appropriate scientific method was followed; that is if the independent variable was systematically changed, the dependent variable was carefully measured and all the other variables were held constant. So an experiment is valid if it was a fair test.

A valid investigation is a fair test. A practical investigation is valid if:

• it is reliable

• it is accurate

• it tests your hypothesis

• it uses appropriate equipment

• it controls the appropriate variables.

Accuracy – (Precise)

The accuracy of an experiment is how precise or correct the results are. This depends on the equipment selected and how carefully the experiment was conducted. Accuracy can be improved by selected more precise equipment, eliminating errors in measurements and calculations and by repeating the experiment several times to calculate an average.

Reliability – (Repeatable)

Results are reliable if repeated experiments lead to consistent results. Sometimes in a large set of repeated measurements, an individual measurement may differ widely from all the others. You may be justified in ignoring this measurement before calculating an average. Measurements cannot be reliable if they are not performed with high precision and accuracy. Measurements can be reliable without being valid. A faulty measuring device, for example, an ammeter not correctly zeroed can consistently provide a false reading which while reliable gives incorrect results.