Hobbies And Interests
Why Bigger is Better
The main reason to use large sample size is to eliminate the chances of a statistical anomaly appearing to be normal, even if you are using a random sample. For example, if you were studying the population of Los Angeles you might randomly selected 100 people; by chance 20 of those people could have red hair. You could draw a conclusion that there are more people with red hair in the city than there really are.
Starting Small
Some people recommend starting out with a small sample sizes. Professor Peter Bachhetti is quoted in The Scientist magazine as saying starting small in certain medical studies can be beneficial because it can determine whether there is value to the study. A large study can take up money and time that may not be justified, he says.
Sampling Factors
Several factors determine the best sample size, starting with the actual size of the population. A sample size should be large enough that it has a small sampling error, which represents how close the results are to your actual population within a percentage. It needs to reflect the diversity in the population, known as the degree of variability. There also needs to be a confidence level such that if the population is repeatedly sampled, the results can be duplicated.
Determining Sample Size
Determining the size of a sample in a study is one of the most difficult tasks for researchers and statisticians. Researchers rely on a number of methods to help them decide the right sample size for them. One way is to take a similar study and use that study's sampling size. Another is through a series of formulas that calculate a sample size using the size of the population and the desired sampling error and confidence level.