Contextual Role of Absolute and Relative Precision in Estimation of Sample Size for Single Proportion in Health Research

Background: Sample size estimation is one of the key components in the initial stage of any health research. The validity of an observational study in estimating prevalence of an outcome of interest is primarily determined by the precision of the estimate. This is generally motivated by the method of sample size estimation. The choice of incorporating absolute precision or relative precision in estimating the sample size of proportions has been a grey area for many years. The objectives of the study were to investigate the role of relative and absolute precision in sample size estimation of proportion and also to provide an easy guide for estimation of sample size using relative and absolute precision using real life examples. Materials and Methods: Sample sizes for different proportions using varying levels of relative and absolute precision were estimated and the variations in the sample size using both methods were graphically plotted. Results: Sample size decreases exponentially with increase in anticipated prevalence in the case of relative precision whereas for absolute precision, it follows a bell-shaped curve. Conclusions: The current study provides scenarios where and how absolute and relative precision can be used. Also, the relation between absolute and relative precision is provided.