In biostatistics, samples differ from population variables due to random fluctuations. The behavior of the random fluctuations can be described by developing mathematical equations. These equations are known as probability distribution functions. The functions give information on how the random fluctuation will exceed an expected level of variation. The distribution functions can be expressed in several ways such as mathematical equations, graphs, and tables. Several distributions can describe random fluctuations. The first example is a normal distribution. The shape of a normal curve is similar to a bell and it is commonly used to analyze several variables that are related to biological studies. This type is most suitable for continuous variables. The second type is the log-normal distribution. In public health, it is commonly used to describe laboratory results and length of hospital stay among other variables. This type of distribution suits continuous variables. The third type of distribution is a binomial distribution.
It describes proportions and it is often used when analyzing the fractions that respond to a specific variable such as treatment of a specific drug that is being analyzed. It is mostly used for discrete random variables. The final type is the Poisson distribution. It describes the number of occurrences of random events. In biostatistics, this distribution function can be used to evaluate the number of deaths that are reported over a specified duration. It is mostly used for discrete variables. Apart from the types described above, several other distributions do not describe the fluctuations that are observed in the data. These distributions often explain the variations in the numbers that are used when testing a specific hypothesis. These distributions are evident when coming up with decisions on whether to reject or not reject the null hypothesis. The most common types of this category of distributions are Chi-square, Fisher F distribution, and Students t distribution (Sullivan, 2011).
All the distributions mentioned above are used in various aspects of biostatistics. However, some are more commonly used than others. The most widely used is the normal probability distribution. This distribution function is described by two parameters. These are the mean and the standard deviation. The distribution describes continuous data that has symmetry. Thus, if a histogram is drawn, then the bars will be evenly distributed on the right and the left from the center. For instance, in public health, a normal distribution function is used when analyzing the birth weights of newborn babies. In most cases, the histogram of weight for newborn babies shows symmetry. Further, a smooth curve can be superimposed on the histogram to prove whether the data follows a normal distribution or not. Thus, by observing the graphs, one can tell whether the data is normally distributed or not (Sullivan, 2011). In the example above, the birth weights follow a normal distribution. The data can be used to estimate a range within which the weights will be considered normal. Observations that lie outside this range are not considered normal. In real life, the World Health Organization has come up with a normal weight range for newborns. If the data provided does not fall in this range, then the babies in the sample are likely to experience the risks that are associated with either low or high birth weights. For instance, if a baby is born with a weight that is below the lower level of this range, then the babyfaces risk of mortality. On the other hand, if the weight of the newborn exceeds the upper limit, then the baby can suffer from obesity. This shows a perfect example of how a normal distribution function is used in public health.
References
Sullivan, L. (2011). Essentials of biostatistics in public health. New York: Jones & Bartlett Publishers.