Parameters are defined as those variables and constants that appear in a mathematical expression, its variation being the one that gives rise to the different solutions to a problem. In this way, a parameter supposes the numerical representation of the enormous amount of information that is derived from the study of a variable. Its calculation is usually carried out through an arithmetic formula that has been previously elaborated from the data obtained from a population.
In the field of computer programming, the use of the term parameter is widely used to refer to an intrinsic property of a procedure.
Why are parameters important?
When a mathematician considers the study of a variable, he has to face a multitude of data that are presented in a disorderly way. That is why a previous work with this information is necessary, reducing and ordering it, in order to work in a simpler and more efficient way.
Although the concentration of the initial data in a parameter entails the loss of part of the information contained in them, this is greatly compensated by being able to make comparisons between the samples or allow a characterization of the data.
Main statistical parameters
Within statistics, three large groups of parameters can be distinguished: position, dispersion and shape.
Position measurements make it possible to identify the value around which the data is mostly grouped. There are two types of dispersion parameters: those with a central tendency (mean, mode and median) and those with a non-central position (percentiles, deciles and quartiles).
For their part, the dispersion measures serve to summarize the distribution of the data. The problem with these parameters is that by themselves they are insufficient by oversimplifying the information, so it is necessary that they be accompanied by other accessory parameters that provide information on the heterogeneity of the data.
Among the most prominent dispersion parameters are variance, standard deviation, coefficients of variation, and range.
Finally, the shape parameters indicate the shape of the histogram of the data, the most common representation being the Gaussian bell. Here it is worth highlighting the coefficients of skewness and kurtosis.
Besides, there are other statistical parameters that are used for a specific purpose, such as the Gini index to measure inequality.
Photos: iStock - mediaphotos / Jovanmandic
