The advantages of evaluating samples from research studies include the ability to apply findings to the general population without having to survey everyone on the planet. Even a survey of just 500 people only has an error margin of +/- 4.5 percent, making it reasonably reliable.
To find out how many people have a fear of public speaking, only asking 10 people would generate a potential error rate of 31.6 percent in comparison with the general population. If a survey yielded seven "yes" responses, the survey would be 70 percent, but the real answer for all people could range between 39 and 100 (really 101) percent. Increasing the sample size up to 500 or 1000 makes the results more reliable.