When a sample is so created that all items or variables are chosen entirely by chance and all items or variables have equal probability of being chosen, the sample is called a random sample.
• Random samples help in getting unbiased results in statistical procedures.
• They provide an easy and inexpensive way to draw conclusions about large samples.
• The only risk attached to a random sample is that even though it is unbiased, it may not be representative of all the variables of a population.
Random samples are used when statistical conclusions have to be drawn about large samples.
In such a case examining all the variables of the population would be a time consuming proposition. Random samples provide small subsets for statistical procedures to be performed on and to derive useful results.
Ten students are to be chosen from a class of fifty students to determine the average height of students of the class. This is done by arbitrarily calling out the roll numbers of students. If there is no biasness whatsoever in choosing the students, the sample of ten students will be a random sample.