Conduct a random sample of a group or population. As an example, suppose you wanted to figure out the average number of push-ups a marine could do after basic training. Rather than test thousands of marines, which would be prohibitive, you would randomly select a small portion and test them. Maybe you choose 100 at random, then your sample size would be 100.
Take measurements from each sample in your sample size. In the example, you would record the number of push-ups each marine could do.
Add up all the values. In the example, say you recorded 100, 150, 125, ...175 for a total of 100 records. You would total all those numbers. Say you summed a total of 15,000 push-ups.
Divide the total by your sample size, i.e., the total number of measurements. In the example, you would divide 15,000 by the sample size of 100, resulting in an average number of push-ups of 150 per marine.