Sample size plays a very crucial role in conducting statistical tests and analyses in various fields, such as business, medicine, nursing, psychology, etc. Sample size is correlated to making an appropriate decision or inference about the product from which the sample has been drawn. In other words, if the sample size is too small, then a systematically conducted study can fail to detect the important effects, associations or correlations. In the same manner, if the sample size is too large, then the study would be complex and tedious. Thus, an optimum sample size is the most important part in any statistical study. This document, therefore, will detail the role played by sample size in business, medicine, nursing, and psychology fields.
In the field of psychology, if one has to compare the difference of means of two populations with a particular sample size, or has to test for a single mean with some sample size, or has to compare the means from two different populations with some sample size using the t-test, then the sample size should be less than 30. If one has to compare the differences of means of two populations with some sample size, or has to test for a single mean with some sample size, or has to compare the means from two different populations with some sample size using the Z- test, then the sample size should be greater than 30. If one has to perform regression analysis in order to predict the attitude of teenagers, generally the sample size should be 10 for each independent variable, defined during that analysis. This means that if there are two independent variables in that study, then the sample size should be at least 20. If, in this study, the data is of categorical type, then the sample size should be more in order to perform the same analysis. In the case of focus group studies, the sample size should be around two from the population and it should have 6-15 groups.
In the field of business, for example in the case of market research studies, the sample size varies from study to study. If one is performing a problem- identification market research study, then the minimum sample size should be around 500 from a population of the size of 1000-2500. If one is performing a problem-solving market research study, then the sample size should be around 200 from a population of the size of 300-500. If one is performing a TV / radio advertising market research study, then the minimum sample size should be around 150, from a population of the size of 200-300. If one wants to test market audits, then the sample size of the stores should be around 10 from a population of 20 stores.
In the field of medicine / nursing, getting a correct sample size is very important. An example can clearly show just how important sample size is in research. In this example, if a researcher wants to know whether or not there is any difference in the curability rate of drug A and drug B, a sample size of 150 patients must be applied to test the two drugs. If in this example there is no difference observed between the two drugs, then a sample size of 150 patients could be the problem. This would be known as a Type II error and this kind of error is very dangerous.
Thus, a larger sample size will avoid this serious error. Sample size, then, plays a crucial role as it is used to avoid Type II errors.
Generally, one should always keep in mind that if the population size is smaller, then one should have a bigger sample size and if the population is large, then one should have a smaller sample size.
For more information on calculating sample size, click here.
