The monthly electricity consumption of 68 consumers: 65-85 (4), 85-105 (5), 105-125 (13), 125-145 (20), 145-165 (14), 165-185 (8), 185-205 (4). Find the median, mean and mode of the data, and compare them.
Hint. All three measures are wanted this time — build the cumulative frequency column once, and it serves the median directly while mode and mean use the frequencies alone.
Finding the median.
Step 1 — Build the cumulative frequency column. 4, 9, 22, 42, 56, 64, 68
Step 2 — Find n/2. n = 68, so n/2 = 34
Step 3 — Find the first class whose cumulative frequency reaches 34. 22 (up to 105-125) is short of 34; 42 (up to 125-145) reaches it. So the median class is 125-145, with cf = 22 (the cumulative frequency of the class before it).
Step 4 — Read off the values and apply the formula. l = 125, cf = 22, f = 20, h = 20 Median = 125 + [(34−22)/20] × 20 = 125 + 12
✦ Median = 137
Mode.
Step 5 — The modal class is the one with the highest frequency: 20, in class 125-145 — the same class as the median, here. l = 125, f₁ = 20, f₀ = 13, f₂ = 14, h = 20 Mode = 125 + [(20−13)/(2×20−13−14)] × 20 = 125 + [7/13] × 20 ≈ 125 + 10.77
✦ Mode ≈ 135.77
Mean.
Step 6 — Class marks: 75, 95, 115, 135, 155, 175, 195. Assumed mean a = 135, h = 20: uᵢ: −3, −2, −1, 0, 1, 2, 3 Σfᵢuᵢ = 4(−3)+5(−2)+13(−1)+20(0)+14(1)+8(2)+4(3) = −12−10−13+0+14+16+12 = 7 mean = 135 + 20(7/68) ≈ 135 + 2.06
✦ Mean ≈ 137.06
Comparison. All three measures land within about 1.3 units of each other (135.77, 137, 137.06) — a sign of a fairly symmetric distribution, with no long tail dragging any one measure far from the others.
Where students slip. Computing all three in isolation without noticing they land close together, missing the chance to catch an arithmetic slip — if one of the three had come out wildly different, that would be the signal to recheck it.
Another way. Check with the empirical relationship: Mode ≈ 3×Median − 2×Mean = 3(137) − 2(137.06) = 411 − 274.12 = 136.88, close to the computed 135.77 — near enough to confirm no major error, given the relationship is approximate.
