The word experiment or random experiment is used for a situation of uncertainty about which we want to have some observations. The actual results of the uncertain situation is called outcome or sample point. In the random experiment, nothing can say with certain about the outcome. An experiment may consist of one or more observations. If there is only a single observation, the term random trail or simply trial is used. A bulb may be selected from a factory to examine if it is defective or not. A single bulb selected is a trial. We can select any number of bulbs. The number of observations will be equal to the number of bulbs. A random experiment has the following properties:
- The experiment can be repeated any number of times. We may select one or more items for inspection. The number of repetitions is called the size of the experiment. In statistics, the size of the random experiment plays a major role in statistical inference.
- A random trial consists of at least two possible outcomes. If a basket contains all the defective bulbs, a selected bulb will be certainly defective. It has only one possible outcome. It is not a random trial. If the basket contains some good and some defective bulbs, a selected bulb will be good or defective. In this case there are two possible outcomes. Thus selecting a bulb from such a basket is a random trial.
- Nothing can be said with certainly about the outcome of the random trial or random experiment. If a sample of four bulbs is selected, may be one bulb is defective. When another sample of four bulbs from the same lot is selected, may be all bulbs are defective. Thus the result of the experiment cannot be predicted even if the experiment is repeated a large number of times.