Mean can be found by Direct Method: x̄ = Σfᵢxᵢ / Σfᵢ. Assumed Mean Method: x̄ = a + (Σfᵢdᵢ / Σfᵢ) where d = x − a. Step Deviation Method: x̄ = a + h(Σfᵢuᵢ / Σfᵢ) where u = d/h.

Median = l + [(n/2 − cf) / f] × h, where l = lower class limit, cf = cumulative frequency before median class, f = frequency of median class, h = class size. Mode = l + [(f₁ − f₀)/(2f₁ − f₀ − f₂)] × h.

A cumulative frequency table shows running totals. Ogive (cumulative frequency curve) is drawn by plotting cumulative frequency against upper class limits. Median can be estimated from the ogive by finding n/2 on the y-axis.

Empirical relationship (approximate): Mode = 3 × Median − 2 × Mean. This relationship holds for moderately skewed distributions. Mean is the most reliable measure but is affected by extreme values (outliers). Median is not affected by outliers and is best for skewed data. Mode is the most frequent value, best for categorical data. For a symmetric distribution, Mean = Median = Mode. In income data, median is preferred as it is not distorted by very high incomes.