Statistics Solutions is the country's leader in statistical consulting and sample size computation. Contact Statistics Solutions today for a free 30-minute consultation.
The sample size computation must be done appropriately because if the sample size is not appropriate for a particular study, then the inference drawn from the sample will not be true and might cause some serious issues.
Suppose the investigator is working on the study of human or animal related subjects. In this case, he needs to utilize the sample size as this will become an essential issue for the sake of moral reasons. The reason for this is because the sample size that is less than the desired number of sample size will expose the subject who is under study to certain harmful treatments because of a lack of knowledge. On the other hand, if the sample size is more than the desired sample size, there will be a necessary number of subjects who are being exposed to a possible harmful treatment or vice versa.
There are various approaches for computing the sample size. The sample size is determined by specifying the preferred width of the confidence interval. There is also a Bayesian approach for sample size determination, which can be used in cases where the researcher wants to optimize the utility function involving the precision of the estimation or the cost. One of the most popular approaches for sample size is that of power.
The researcher should keep in mind that the sample size requires both the technical skills of a statistics professional and the scientific knowledge of a researcher. Sample size is determinable from a type of cost/benefit analysis. This is because the sample size is related to the cost of an experiment and the sample size is also often directly related to the cost saving during the improvising of the process.
Usually, the study on which the researcher works is often based on a limited budget, so this affects the sample size. An alternative way to get rid of the sample size problem is to make the sample size fixed for certain studies. But this way of keeping the sample size fixed is also not useful when the researcher wants to widen his scope of study in terms of additional suppliers of raw materials, broader demographics of the subjects, etc.
The researcher should keep in mind that there are different types of sample size problems. The sample size problem involving moral issues in an opinion poll is very different from those which involve medical experiments. Also the outcomes of the usage of a sample size more than the desired sample size and the usage of the sample size less than the desired sample size is not the same.
Sample size is generally more crucial in cases that take a huge amount of time while performing data collection.
