How can you insure a test is fair?

It happens at various instants of our life that we need to analyze something. For example, you need to determine a better brand of cookies from three or four available brands. Similarly, your television was all okay yesterday, but it is not showing any channel today. So, in such circumstances, you will need to analyze what is creating the problem and what can solve the problem. This analyzing is also known as 'testing'. This everyday testing is derived from the methods employed by scientists in evaluating their ideas and thoughts. A fair test can help you determine the actual problem and its best solution as well. There are few factors or pillars of testing which should be considered in order to have a fair test.

Comparison is the first and most important factor in having a fair test. For example, you are going to test why your car is not getting started. In this case, you may suspect a number of problems. For example, you think some of the wires are not plugged in correctly. You should try to start the car without plugging in those wires so that you can derive that those wires have 'something' do with your car starting. Similarly, you should be capable of comparing results in order to judge your testing. After 'comparison', studying employed variables is of immense importance. For example, in the case of a car, you should be confident that these are the wires which are not making the car start and not the fuel that has run off. In simple words, control variable are all those 'possibilities' which can cause a problem. You should also be confident on the effect of each cause.

The Value of Fairness in Testings

Bias is a factor that can damage the fairness in your testing. For example, you want to have a fair test in order to determine the better and tastier brand of cookie. Now, your tester may not be able to give you right judgments if he/she is hungry. In this case, the tester may like the first batch of cookies and not the second one. Similarly, the tester may like a specific brand and cannot judge that the other brand can be even tastier. In such circumstances, it is recommended to pack batch of cookies in envelopes instead of the original brand packet. You can name the envelopes as Cookie A and Cookie B. This will give more fair testing results.

One of the most important factors to have fairness in testing is the capability of distinguishing between the random and real causes. In case of biscuits, the oven may have burnt some cookies placed at the deeper side of the oven. This can cause a change in taste. Similarly, the uneven burning of the oven can also affect the test results. All of these factors are called as 'random factors' which can affect the outcomes of your testing. Finally when you have rated biscuits in hand and you see a minor difference between the ratings of the two batches of biscuits, you should be capable of determining whether this change in rating is due to random factors or due to real factors. Sample size is a very important thing to consider in order avoiding the involvement of random factors. Similarly, results should be taken from various biscuit testers and not only from three or four people. This will solidify any fair test results.

More Information On Fairness in Testing

1. FairTest
2. Frontline: Secrets of the SAT
3. Rights and Responsibilities of Test Takers
4. What Do Test Scores in Texas Tell Us?