Sample Size Computations

How many units of observations do you need to draw data from? This is normally a neglected but very important question during the design of a study. We will help you arrive at appropriate sample sizes in most situations, depending on your study design and hypotheses.

Determining sample size is very imperative; samples that are too large may waste time, resources and money and at worst, could make the conduct of the study infeasible. Samples that are too small on the other hand may lead to imprecise results. In many situations the minimum sample size needed to estimate a process parameter, such as the population mean or proportion, can easily be determined. Sample size problems are context-dependent moreover it is only one aspect of a quality study design and should not outweigh other factors that characterize a good research design.

Larger sample sizes principally result in increased precision when estimating unknown parameters. In practice, the sample size used in a study has to balance between the expense of data collection and processing, and the need to have sufficient statistical power to detect important differences.

Perhaps one of the most frequent questions asked of a statistician is, "How many units should be included in the sample?" It is unfortunate that this question is always met with more questions instead. There is no correct answer without additional information or assumptions. For instance, the appropriate sample size for a population-based survey is determined largely by three factors: (i) the estimated prevalence of the variable of interest (ii) the desired level of confidence and (iii) the acceptable margin of error. These are the kind of questions that the statistician would expect answers to, from the investigator/researcher, in order to proceed and calculate the required minimum sample size.

In case of sample size required for a study involving estimating and/or comparing unknown population means, the final number of sampling units required will be influenced by: (i) value selected for the risk of rejecting a true hypothesis (what is conventionally referred to type I error), (ii) the risk of accepting a false null hypothesis when a particular value of the alternative hypothesis is true (what is referred to as type II error) and (iii) the value of the population standard deviation.

Successful resolution of the sample size requires a close and honest and collaboration between statisticians and subject matter experts. We are happy to help you arrive at appropriate sample sizes for most situations, depending on your study design and hypotheses.

Please contact us now for a personal attention to your queries.