Introduction
Decile is an analytical method used in statistical studies to divide a set of data into ten equal subsections. The ranking is either in an ascending order or in descending order. The division is similar to the division done using percentiles or quartiles.
Understanding Decile
Decile facilitates categorisation of data sets and observations into samples for convenience of analysis and measurement. The data set is divided into ten equal parts and will contain nine data points.
The purpose of deciles is to determine the largest and smallest values based on a metric. For example, the S&P BSE 100 Index (Indian stock exchange) can be split into ten deciles (10 firms in each decile) on the P/E multiple of firms to determine and analyse the highest and the lowest P/E firms comprised in the index.
After the division of the data in deciles, the data set is assigned decile ranks. Decile ranks enable arrangement of the data from the lowest value to the highest value on a scale of one to ten. Each rank is based on an increase of ten percentage points. D1-the first decile has 10 per cent of the data set below it, D2-the second decile has 20 per cent of the data set below it, D3-the third decile has 30 per cent of the data set below it, and similarly for the rest of the data.
The formula for calculation of a decile is mentioned below: D1 = Value of [n+1/10]th data D2 = Value of [2(n+1)/10]th data D3 = Value of [3(n+1)/10]th data D9 = Value of [9*(n+1)/10]th data
The 5th decile is the median. The data point D5 = Value of [5*(n+1)/10] is the data point half-way through the distribution.
Deciles are used in finance and economics to divide data into sets for the purpose of analysis. For example, the data set of mutual fund portfolio or the data set of income-tax return filers can be divided for analysing the top 10 per cent and so on.
Conclusion
Deciles as statistical tools can be used for a variety of purposes and in a variety of industries or segments. The government, too, can use the data to frame welfare policies, tax policies, and so on.
