Difference Between Descriptive and Inferential Statistics
In the study of statistics, inferential and descriptive statistics are two important areas that need to be understood well. Very many people use these two terms interchangeably. Strangely, the goals and methodologies used are a bit different though the statistical measures are the same.
Description Statistics
To a layman, descriptive statistics are used when relating to a sample. It is that simple and straightforward. It is used when trying to understand the various observations derived from a sample. From a population given, you take a sample, record the data and summarize the various properties of the sample. Uncertainty is eliminated because only the items measured are described. No relationship is made with the larger population from which the sample was derived. The following statistical measures are used in the description of groups under descriptive statistics: central tendency (mean, mode and medium), dispersion (measures how far from the center the data is extended) and skewness (tells the user how symmetrical the values are). In descriptive statistics, graphs are frequently used.
Inferential Statistics
Inferential StatisticsInferential statistics uses the sample to make conclusions about the population from which the sample was derived. Since the main objective of inferential statistics is to test a sample and to use the findings thereon to generalize the whole population, it is important that we demonstrate confidence that the sample used is a true representation of the population. Broadly, we must define the whole population that is under our study, draw a sample that is representative of the population and finally, analyze the likely sampling error. The common analytical tools used in inferential statistics are regression analysis, confidence intervals and hypothesis tests.
Difference Between Descriptive and Inferential Statistics
As discussed above, the main difference between descriptive and inferential statistics is in the method or process as it is in the conclusions that are drawn from the various analytical tools used. When looking at descriptive statistics, we are required to choose that group we want to describe. We then measure the various subjects therein. The statistical conclusion will be described in this group with almost complete certainty since the measurement error has already been stated. In the case of inferential statistics, all that is needed is to define the population. We then proceed to devise a sampling methodology or plan that will produce a sample that is representative.
The statistical results obtained will incorporate the inherent uncertainty. There is an inherent risk of using the sample obtained to try and understand the whole population that was stated. The last difference between descriptive and inferential statistics is that it is simpler to perform descriptive statistics. Care should, however, be taken to ensure that any assumptions/errors of margin are likely computed and documented. On the other hand, if a more accurate and representative relationship is to be obtained between the sample and population, then inferential statistics is the ideal method to be used. Inferential statistics will set the hypotheses that are to be tested, and the confidence levels are clearly defined using the population parameters.
