High Performance Statistical Queries: Linear Dependencies Between Continuous Variables

high performance statistical queries, covariance, correlation coefficient, coefficient of determination, linear regression, linear dependence, sql dependencies

In my previous articles, I dealt with analyses of only a single variable. Now it is time to check whether two variables of interest are independent or somehow related. For example, a person’s height positively correlates with shoe size. Taller people have larger shoe sizes, and shorter people have smaller shoe sizes. You can find this and many more examples of positive associations at: http://examples.yourdictionary.com/positive-correlation-examples.html. A negative association is also possible. For example, an increase in the speed at which

Continue Reading

High Performance Statistical Queries –Skewness and Kurtosis

In descriptive statistics, the first four population moments include center, spread, skewness, and kurtosis or peakedness of a distribution. In this article, I am explaining the third and fourth population moments, the skewness and the kurtosis, and how to calculate them. Mean uses the values on the first degree in the calculation; therefore, it is the first population moment. Standard deviation uses the squared values and is therefore the second population moment. Skewness is the third, and kurtosis is the

Continue Reading

High Performance Statistical Queries in SQL: Part 3 – Measuring the Spread of a Distribution

Besides knowing the centers of a distribution in your data, you need to know how varied the observations are. In this article, we’ll explain how to find the spread of a distribution. Are you dealing with a very uniform or a very spread population? To really understand what the numbers are saying, you must know the answer to this question. In the second part of this series, we discussed how to calculate centers of distribution. Like the center, the spread can

Continue Reading

High-Performance Statistical Queries in SQL: Part 2 – Calculating Centers of Distribution

My previous article explained how to calculate frequencies using T-SQL queries. Frequencies are used to analyze the distribution of discrete variables. Today, we’ll continue learning about statistics and SQL. In particular, we’ll focus on calculating centers of distribution. In statistics, certain measurements are known as moments. You can describe continuous variables (i.e. a variable that has a large range of possible numbers, such as household incomes in a country) with population moments. These moments give you insight into the distribution

Continue Reading

High-Performance Statistical Queries in SQL: Part 1 – Calculating Frequencies and Histograms

Database and Business Intelligence (BI) developers create huge numbers of reports on a daily basis. Many of these reports include statistical analyses. How can you perform statistical queries in SQL? Statistics are very useful as an initial stage of a more in-depth analysis, i.e. for data overview and data quality assessment. However, there are not many statistical functions in SQL Server. In addition, a good understanding of statistics is not very common among T-SQL practitioners. In SQL Server 2016, you

Continue Reading