Perlukah menghitung ukuran sampel?

Abstract

Sample size is an issue worth-considering but not the most essential thing to fulfil for a good research. A much more crucial cause of concern to any research is the validity of inference a research is drawing, i.e. the extent to which the research is able to control systematic error that stems from bias and confounding. Sample size refers to random error; it does not address nor correct systematic error. The larger sample size, the less random error, the more precise estimates a research can yield about difference/ association/ effect of a variable(s). Most of the assignment of values in any sample size formula is arbitrary. As such, the product of estimating sample size, regardless of the formula being used,  cannot be viewed as an absolute end; the actual sample size can be larger or smaller than the estimated one. Beyond statistical aspect, several other important factors should be considered when estimating sample size, such as ethics, cost, and the amount oftime available for conducting the research.