Statistics · Ch 3 — Measures of Central Tendency
Geometric Mean — Computation, Merits and Limitations
Geometric Mean — Computation, Merits and Limitations
Geometric mean (GM) is the root of the product of positive observations. It is the correct average to use whenever data involves rates of change, ratios, growth rates or index numbers — situations where values multiply together rather than simply add.
1. Individual series:
For more than two or three values, this is computed conveniently using logarithms:
2. Discrete series (using logarithms):
3. Continuous series — use the class mid-point in place of :
A common business use: the average rate of growth over several years must always be found by geometric mean, never by arithmetic mean — e.g. if a firm's sales grow by different percentages in different years, only the GM of the year-on-year growth FACTORS gives the true average annual growth rate that, compounded over the same number of years, reproduces the actual overall growth.
Merits of geometric mean:
- Based on all observations.
- Rigidly defined, capable of further algebraic treatment.
- The most suitable average for ratios, rates of change, and growth rates — gives less weight to large extreme values than the arithmetic mean does.
- Useful in the construction of index numbers (covered in Std 12 Statistics).
Limitations of geometric mean: …
The nth root of the product of n positive observations; computed in practice via logarithms as the antilog of the average of the logarit …
A value of the form (1 + rate), used so that the geometric mean of successive years' growth factors gives the correct compound …