What is the difference between fair, valid and reliable?

When conducting any test or experiment, it is ideal for it to be fair, valid and reliable.

A fair test is one where one variable is changed at a time, for testing its particular effect on the experiment, while keeping all other variables constant. For a test to be fair, a control group must be implemented, biases must be limited, an adequate sample size must be used, and interfering variables have to be controlled. Fairness ensures credibility of the experiment itself.

A reliable test is one in which there are many trials, and each time the test is run, the results are consistent with each other. Reliability ensures accuracy and precision of data.

A validity test is what it sounds like--it measures the validity of the test in conveying what it claims to show. A valid test is said to have reasonable controlled variables and have a proper method and procedure that truly measures what the experiment was meant to measure. Validity ensures that the data is believable in that it accomplishes what the test aimed to accomplish by reasonable means.

When a test is fair, valid, and reliable, it can be considered credible, as the data is both accurate and precise in forming its conclusion(s) and the way in which data was collected is justified and reasonable.