Meaning, Objectives and Types of Dispersion

An average — the mean, median or mode studied in the previous chapter — condenses an entire data set into a single representative number, but it says nothing about how the individual values are actually spread around that number. Two shops could both report an average daily sale of ₹5,000: one whose daily sales stay within a narrow band of ₹4,800–₹5,200 every day, and another whose sales swing wildly between ₹1,000 and ₹9,000. The two averages are identical, yet the two businesses are nothing alike in terms of predictability and risk. Dispersion (also called variability or scatter) measures exactly this — the extent to which individual observations differ from one another and from their central value.

Objectives of studying dispersion — (i) it tests how reliable an average is: a small dispersion means the average is a good, dependable summary of the data, while a large dispersion means the average alone can be misleading; (ii) it allows two or more series to be compared for consistency or stability, which matters directly in business — comparing the steadiness of two salesmen's monthly targets achieved, or two machines' hourly output; (iii) it is the essential foundation for more advanced statistical techniques such as correlation, regression and quality control, all of which are built on measuring variability; and (iv) it helps identify and control irregularity — a firm that understands how variable its costs or output are is better placed to plan and budget for it.

Absolute measures of dispersion are expressed in the same unit as the original data (rupees, kilograms, marks, days) — Range, Quartile Deviation, Mean Deviation and Standard Deviation all belong to this group. They are ideal for describing the spread of a single series, but they cannot fairly compare two series that are in different units, or even the same unit but very different average sizes. Relative measures of dispersion (also called coefficients) remove this limitation by expressing the absolute measure as a ratio — usually of the corresponding measure of central tendency — producing a pure, unit-free number that can be compared directly across any two series, however different their scale. Every absolute measure in this chapter has its own matching coefficient, and this chapter builds both together, measure by measure, exactly as prescribed in the Odisha CHSE (Council of Higher Secondary Education) +2 first year Commerce Business Mathematics and Statistics syllabus — the same statistical principles that measures of dispersion are built on in commerce and statistics curricula across the country.