Measures of Central Tendency
Condensing an entire data set into one representative 'average' value using the arithmetic mean, median and mode.
It powers CSAT (Prelims Paper-II) data-interpretation and basic-numeracy questions on mean, median and mode, and decides how you read every economic statistic — per capita income, NSS averages, inflation indices. For GS-III economy it arms you to argue why a bare average can hide income inequality and why the median is often the more honest welfare indicator. Such averages are the daily tools of India's nodal statistical body, MoSPI (CSO/NSSO).
Understand the chapter
The Idea: One Number for the Whole Data
Central tendency is a numerical method that summarises a large data set in a single representative or 'typical' value standing in for the entire data. The chapter frames it through Baiju, a small farmer in Balapur village (Buxar district, Bihar), whose 1-acre holding is judged against the village's 50 farmers using three lenses — the mean (above average in the ordinary sense), the median (above what half the farmers own) and the mode (above what most farmers own).
- Three commonly used averages: Arithmetic Mean, Median, Mode.
- Two more exist for special cases: Geometric Mean and Harmonic Mean (not computed in this chapter).
- A good average is a typical value that represents the whole distribution.
Arithmetic Mean — the Workhorse Average
The arithmetic mean is the most commonly used average: the sum of all observations divided by their number, denoted X̄ (X-bar). Symbolically X̄ = ΣX/N. It is intuitive and uses every value, which is also its weakness — a single extreme value drags it up or down.
- Formula: X̄ = ΣX/N, where ΣX = sum of observations, N = number of observations.
- Example: marks 40, 50, 55, 78, 58 give a mean of 56.2.
- Computed for ungrouped data and grouped data (discrete & continuous series).
Three Ways to Compute the Mean
For small data the Direct Method (ΣX/N) suffices. When figures are large, the Assumed Mean Method picks any value A, takes deviations d = X − A, and gives X̄ = A + Σd/N. The Step-Deviation Method further divides deviations by a common factor c, giving X̄ = A + (Σd'/N) × c — the last step, multiplying back by c, is the one aspirants forget.
- Discrete series (direct): X̄ = ΣfX/Σf.
- Continuous series: replace each class by its mid-point, then proceed as in discrete series.
- Assumed mean A can be any value, ideally a centrally located one to ease arithmetic.
Properties of the Mean & the Weighted Mean
Two properties are exam favourites: the algebraic sum of deviations of items about the mean is always zero, Σ(X − X̄) = 0; and the mean is sensitive to extreme values. When items differ in importance, a Weighted Arithmetic Mean assigns weights W, giving ΣWX/ΣW — the conceptual seed of Index Numbers like CPI/WPI.
- Σ(X − X̄) = 0 — deviations cancel out.
- Mean is pulled by outliers; replace 12 by 96 in a series and the mean jumps.
- Weighted mean = ΣWX/ΣW; weights reflect relative importance (e.g., budget shares).
Median — the Positional Middle
The median is the middle value when data are arranged in order of magnitude, splitting the distribution into two equal halves. Crucially, it is a positional average and is not affected by extreme values — the size of the largest item can rise without moving it. Its position is the (N+1)/2 th item; for an even count it is the mean of the two middle values.
- Position of median = (N+1)/2 th item (individual & discrete series, located via cumulative frequency).
- Continuous series: first find the median class where the N/2 th item lies (note: N/2, not (N+1)/2).
- Formula: Median = L + ((N/2 − c.f.)/f) × h.
- L = lower limit of median class, c.f. = c.f. of preceding class, f = frequency of median class, h = class width.
Quartiles & Percentiles — Slicing the Distribution
Quartiles divide ordered data into four equal parts. Q1 (lower) has 25% of items below it, Q2 is the median (50% below), and Q3 (upper) has 75% below it; Q1 and Q3 bracket the central 50% of the data. Percentiles cut the data into 100 parts with 99 dividing points P1–P99, where P50 equals the median.
- Q1 = size of (N+1)/4 th item; Q3 = size of 3(N+1)/4 th item.
- Central 50% of data lies between Q1 and Q3.
- P50 = median; scoring the 82nd percentile means you stand below 18% of candidates.
Mode — the Most Typical Value
The mode is the value that occurs most frequently — the 'most typical' observation, and the lens behind 'above what most farmers own' in the Baiju example. It is the average that best answers 'what is the commonest size or value?', useful for categorical and most-frequent data.
- Mode = the most frequently occurring (most typical) value in a series.
- Best for questions about the commonest category (e.g., modal plot size or shoe size).
Key terms
- Central Tendency
- A single representative value that summarises an entire data set.
- Arithmetic Mean (X̄)
- Sum of all observations divided by their number: X̄ = ΣX/N.
- Assumed Mean Method
- Mean via an assumed value A: X̄ = A + Σd/N, where d = X − A.
- Step-Deviation Method
- Mean using deviations scaled by a common factor c: X̄ = A + (Σd'/N) × c.
- Weighted Arithmetic Mean
- Average that weights items by importance: ΣWX/ΣW.
- Median
- The middle positional value dividing ordered data into two equal halves.
- Median Class
- In a continuous series, the class interval containing the N/2 th item.
- Quartiles
- Values dividing ordered data into four equal parts (Q1, Q2, Q3).
- Percentiles
- Values dividing data into 100 equal parts (P1–P99); P50 = median.
- Mode
- The most frequently occurring / most typical value in a series.
Must-know facts exam-ready
- Three common averages: Arithmetic Mean, Median, Mode; two special ones: Geometric Mean and Harmonic Mean.
- Arithmetic Mean = ΣX/N, denoted X̄ (X-bar).
- Assumed Mean Method: X̄ = A + Σd/N, where d = X − A.
- Step-Deviation Method: X̄ = A + (Σd'/N) × c, where d' = (X − A)/c.
- Discrete-series mean: X̄ = ΣfX/Σf; continuous series uses class mid-points.
- Property: Σ(X − X̄) = 0; and the mean is affected by extreme values.
- Median position = (N+1)/2 th item (individual/discrete series).
- Continuous-series median locates the N/2 th item, NOT (N+1)/2.
- Continuous median formula: L + ((N/2 − c.f.)/f) × h.
- Quartiles: Q1 = (N+1)/4 th item, Q3 = 3(N+1)/4 th item; Q1–Q3 hold the central 50%.
- Percentiles: 99 dividing points P1–P99; P50 = median.
- Median and mode are unaffected by outliers; the mean is not.
Memory tricks remember it for good
Traps to avoid
- Reversing sensitivity: the MEAN is affected by extreme values; the MEDIAN (and mode) are NOT — UPSC flips this.
- Continuous-series median uses N/2 to find the median class, but individual & discrete series use (N+1)/2; mixing them is the classic error.
- In the median formula, c.f. is the cumulative frequency of the PRECEDING class, never the median class itself.
- Percentile reading: the 82nd percentile means 82% lie below you (you are below the top 18%), not that you 'scored 82'.
- Step-deviation answers go wrong if you forget to multiply the deviation term back by the common factor c.
- Weighted mean divides by ΣW (sum of weights), not by N — it is not a simple average.
Exam focus
🧠 Prelims angles
- CSAT Paper-II: directly compute mean, median or mode from a small data set.
- Conceptual MCQ: which average suits skewed data (median) vs commonest value (mode) vs overall total (mean).
- Formula recall: median position (N+1)/2, quartiles (N+1)/4 and 3(N+1)/4, P50 = median.
- Identify the median class (N/2) and apply the interpolation formula in a continuous series.
- Effect of outliers: how one extreme value shifts the mean but not the median.
- Weighted average and its link to Index Numbers (CPI weights, per capita income).
✍️ Mains angles GS-III
- Why per capita income (a mean) can mislead on welfare, and why median income is a more honest gauge.Contrast mean vs median on a right-skewed income distribution; connect to inequality and the 'average Indian' fallacy.
- Choosing the right average is a policy decision, not a formality.Map mean→income/production totals, median→typical income, mode→commonest land-holding; argue context dictates the measure.
- Weighted averages as the backbone of index numbers (CPI/WPI, inflation).Show how commodity weights by importance turn a simple mean into a policy-relevant price index.
Last-minute revision tick as you recall
- Central tendency = one representative value for the whole data.
- Three averages: Mean, Median, Mode (plus GM & HM for special cases).
- Mean = ΣX/N; uses all values; pulled by outliers; Σ(X − X̄) = 0.
- Mean methods: Direct, Assumed (A+Σd/N), Step-deviation (A+(Σd'/N)×c).
- Median = ordered middle; position (N+1)/2; robust to extremes.
- Continuous median: locate N/2 class → L + ((N/2 − c.f.)/f) × h.
- Quartiles Q1/Q2/Q3; central 50% lies in Q1–Q3; P50 = median.
- Mode = most frequently occurring / most typical value.
- Baiju, Balapur (Buxar, Bihar): mean=above average, median=above half, mode=above most.
Distilled from NCERT Class 11 · Statistics for Economics for UPSC. Always cross-check facts with the original NCERT.