Inferential statistics is a branch of statistics.
Statistics is the area of mathematics that works with large groups of numerical data to, based on the calculation of probabilities, generate inferences : that is, conclusions that are deduced from something else. The analysis of quantitative information of a human or natural expression is also called statistics.
Inferential statistics is the division of statistics focused on the processes and techniques that make it possible to know the characteristics of a statistical population by studying a part of it . In this way, by examining the data provided by a sample, deduction is used to obtain a conclusion about the total of the set.
It can be said that inferential statistics uses procedures and models to draw inferences about the population in question. Its methods are varied and allow multiple objectives to be satisfied.
Concepts linked to inferential statistics
To understand what inferential statistics is, it is important to be clear about several notions. The idea of statistical population , for example, refers to a set of events or elements that, sharing certain traits or properties , are of interest for the development of an experiment or for answering a question.
The population can be made up of a precise and limited number of components or be potentially infinite. Inferential statistics, in this framework, works with a sample to generate its inferences.
A sample is a subset of the population. This subset needs to be representative of the total sample, for which the appropriate technique must be used for its determination.
With the sample already defined, inferential statistics examines the variables that enable the representation of a phenomenon. Statistical variables, as their name indicates, can vary, with this variability being represented through different numerical values.
Inferential statistics works with a sample to generate inferences about a population.
Procedures and conclusions
Inferential statistics, in short, is an appropriate resource for analyzing a population , producing knowledge about some of its properties or trends . Starting from a sample, comparisons, interpretations and projections can be made.
The deductions made by inferential statistics can be carried out with parametric evaluations, time series, correlation tests or hypothesis tests, for example. Thus, what is observed in the sample (subset) can be “transferred” to the population (set).
Elections and inferential statistics
Electoral polls are usually developed according to inferential statistics. It must be taken into account that all voters constitute the statistical population; For reasons of practicality, since it is generally not possible to survey that many people , a sample must first be selected.
It is important that the members of the sample exhibit different socioeconomic characteristics, so that they reflect the diversity of the population. The sample can be formed with inhabitants of different cities or different neighborhoods of a locality, to mention two possibilities.
Once the survey has been carried out on the sample, asking the participants which candidate they plan to vote for, the percentages that reveal, numerically, the voting intention can be calculated. By deduction, finally, it indicates which candidates are the preferred ones at a general level.