Geometric mean and harmonic mean
Business Statistics · Unit II · Topic 6 of 15 · 4 min read
On this page
Cases checked to exist. Read the judgment before quoting a holding.
Every case cited here has been checked to exist — the citation resolves and names the right parties. The propositions attributed to each case have not been verified line by line against the judgment. Case names link through to the full judgment: read it before quoting a holding in an examination answer, and never carry a proposition from notes into court on trust.
1. Geometric mean
The geometric mean of n positive values is the nth root of their product: GM = (x1 × x2 × … × xn)1/n. Taking logarithms:
- Individual: GM = antilog (Σlog x / n).
- Discrete or continuous (x = mid-point): GM = antilog (Σf log x / N).
Uses: averaging rates of change, ratios and percentages (growth of population, rates of interest, depreciation); index numbers (Fisher’s index is the GM of Laspeyres and Paasche). Example: sales rise by 10% in year 1 and 40% in year 2. The average growth factor is √(1.10 × 1.40) = √1.54 = 1.241, i.e. 24.1% a year, not the arithmetic 25%.
Merits: rigidly defined; based on all items; gives less weight to large items than the AM; suitable for rates and ratios; capable of algebraic treatment. Demerits: hard to compute without logs; cannot be found if any value is zero or negative; may be a value not in the series.
2. Harmonic mean
The harmonic mean is the reciprocal of the arithmetic mean of the reciprocals:
- Individual: HM = n / Σ(1/x).
- Discrete or continuous: HM = N / Σ(f/x).
Uses: averaging speeds, rates and prices where the numerator is fixed (distance fixed, time varies; money fixed, quantity varies). Example: a car goes 60 km at 60 km/h and returns at 40 km/h. Average speed = 2 / (1/60 + 1/40) = 2 / (5/120) = 48 km/h, not 50. (Check: 1 hour + 1.5 hours = 2.5 hours for 120 km = 48 km/h.)
Merits: rigidly defined; based on all items; gives the right answer for rates; gives most weight to small items. Demerits: not easily understood; cannot be found if any value is zero; gives too much weight to small values for most other purposes.
Relation: for any set of positive, unequal values, AM > GM > HM (equal only when all values are equal). For two values, GM = √(AM × HM).
3. Worked: harmonic mean, discrete (January 2026)
| x | f | 1/x | f/x |
|---|---|---|---|
| 124 | 5 | 0.008065 | 0.040323 |
| 129 | 18 | 0.007752 | 0.139535 |
| 134 | 20 | 0.007463 | 0.149254 |
| 139 | 7 | 0.007194 | 0.050360 |
| 144 | 3 | 0.006944 | 0.020833 |
| Total | 53 | 0.400305 |
HM = 53 / 0.400305 = 132.40. (For comparison, AM = 7,027/53 = 132.58; the HM is a little lower, as it must be.)
4. Worked: geometric mean, grouped (January 2026)
| Class | f | m | log m | f log m |
|---|---|---|---|---|
| 0–10 | 5 | 5 | 0.6990 | 3.4950 |
| 10–20 | 7 | 15 | 1.1761 | 8.2327 |
| 20–30 | 15 | 25 | 1.3979 | 20.9685 |
| 30–40 | 25 | 35 | 1.5441 | 38.6025 |
| 40–50 | 8 | 45 | 1.6532 | 13.2256 |
| Total | 60 | 84.5243 |
Σf log m / N = 84.5243 / 60 = 1.4087. GM = antilog 1.4087 = 25.63. (AM = 29.0, so GM < AM, as expected.)
5. Worked: both GM and HM (June 2025)
| Class | f | m | log m | f log m | f/m |
|---|---|---|---|---|---|
| 100–200 | 4 | 150 | 2.1761 | 8.7044 | 0.026667 |
| 200–300 | 6 | 250 | 2.3979 | 14.3874 | 0.024000 |
| 300–400 | 10 | 350 | 2.5441 | 25.4410 | 0.028571 |
| 400–500 | 12 | 450 | 2.6532 | 31.8384 | 0.026667 |
| 500–600 | 11 | 550 | 2.7404 | 30.1444 | 0.020000 |
| 600–700 | 13 | 650 | 2.8129 | 36.5677 | 0.020000 |
| Total | 56 | 147.0833 | 0.145905 |
GM = antilog (147.0833 / 56) = antilog 2.6265 = 423.1.
HM = 56 / 0.145905 = 383.8.
Check: AM = 25,500 / 56 = 455.4 > GM 423.1 > HM 383.8.
August 2024 set the same pair on marks 0–10 to 50–60 (f = 7, 12, 4, 25, 30, 22). Follow the same layout with mid-points 5, 15 … 55.
How this is examined
HM from a discrete table and GM from a grouped table: January 2026 (six marks each, as alternatives). GM and HM from one table: June 2025 and August 2024 (ten marks). A theory answer on “measures of central tendency” (January 2025, August 2024) should include a paragraph each on GM and HM, their uses and the relation AM ≥ GM ≥ HM.
Now try the real questions: Business Statistics past question papers.