Statistics · Ch 3 — Measures of Central Tendency
Relationship Among AM, GM and HM; Choosing the Right Average
Relationship Among AM, GM and HM; Choosing the Right Average
For ANY set of two or more positive observations that are not all equal, the three means always stand in the following fixed order:
with equality holding in all three () only when every observation in the data set is identical. This relationship is a useful, quick arithmetic cross-check: after computing all three means for the same data, verifying that they fall in this order (or are exactly equal for uniform data) confirms none of the three has been mis-computed.
Special case (exactly two observations): for positive values and :
This identity holds ONLY for two observations — it is not a general relationship for values, a distinction the Gujarat Std 11 Commerce Statistics syllabus expects a student to state precisely rather than over-generalise.
Choosing the appropriate average — a summary table:
| Situation | Best average | Why |
|---|---|---|
| General-purpose summary, further statistical work (dispersion, correlation) needed | Arithmetic mean | Based on all values; algebraically tractable |
| Data has extreme values / income-type skewed distribution | Median | Unaffected by outliers |
| Need the single most "typical" or most commonly occurring value | Mode | Represents the most frequent case directly |
| Averaging ratios, rates of change, growth rates, index numbers | Geometric mean | Correctly averages multiplicative/compounding data |
For any set of positive, non-identical observations, the arithmetic mean is always the largest of the three means, the harmonic mean the smallest, and the geometric mean always lies between them; all three are equal on …