EMJ series on statistics and methods: normal distribution and the central limit theorem

ABSTRACT

This second article in the EMJ Series on Statistics and Methods dwells on the basics of sampling distribution of' variables', which are presented in detail in the preceding article in this Issue of the Ethiopian Medical Journal (EMJ). The present article highlights recommended routines that need to be undertaken in order to understand information collected in a particular study before embarking on doing complex statistical analyses. It underscores the importance of descriptive statistics as a means to getting insights into data quality and learn about the scale and distribution of different variables in a data set. The article emphasizes the need for assessing the sampling distributions of variables as a prerequisite to making decide on selection of appropriate statistical techniques for in a data set. It describes salient features of a normally distributed random variable and touches on some other probability distributions commonly used in epidemiological studies. The article also describes the central limit theorem highlighting salient points on its conceptual basis in understanding sampling distributions of sample means and the implications of using normal distribution to make inference about the population based on summary measures from a sufficiently large sample