Chapter 5 — Measures Of Central Tendency

What are measures of central tendency?

Measures of central tendency are statistical values that describe a central or typical value for a probability distribution or a data set. They summarize a large set of observations by a single representative number that indicates where the middle or center of the data lies.

Why they are important (Economics context):

• Help summarize incomes, prices, production, exam scores, etc., into a single representative value.
• Useful for policy decisions, comparisons between groups, and identifying typical behavior.

Main measures:

• Arithmetic Mean (Average): Sum of all observations divided by the number of observations. It uses every value and is useful when values are quantitative and not heavily skewed.
• Median: The middle value when observations are arranged in ascending (or descending) order. It divides the distribution into two equal parts and is resistant to extreme values (outliers).
• Mode: The most frequently occurring value(s) in the data. Useful for categorical data and to identify common/typical categories.

Interpretation and selection:

• If the distribution is symmetric and has no extreme values, the mean is a good representative; mean ≈ median ≈ mode.
• For skewed distributions (e.g., incomes), the median often gives a better idea of a "typical" value because it is not affected by extremes.
• For categorical or discrete data where frequency matters (e.g., most preferred product size), the mode is informative.

Properties (brief):

• Mean uses all observations; influenced by extreme values.
• Median depends only on order; unaffected by magnitude of extremes.
• Mode depends on frequency; a distribution can be unimodal, bimodal or multimodal.

Limitations:

• Mean is not suitable for highly skewed data or when data contain outliers.
• Median ignores values except their order; not useful for algebraic manipulation like sum of deviations = 0 property of mean.
• Mode may be non-unique or unstable for small samples.

Empirical relation (for moderately skewed, unimodal distributions): Mode ≈ 3 × Median − 2 × Mean (approximate).

How to compute (overview):

1. Mean (ungrouped data): Add all observations and divide by n.
2. Mean (grouped data): Use class mid-points multiplied by class frequencies, then divide by total frequency.
3. Median (ungrouped): Sort data; if n odd take (n+1)/2 th value, if n even average the n/2 and (n/2 + 1)th values.
4. Median (grouped): Locate median class (cumulative frequency ≥ N/2) and apply the interpolation formula for class intervals.
5. Mode (ungrouped): The value with highest frequency. For grouped data, find modal class and apply grouped-data mode formula (interpolation).

These measures are the foundation for many other statistical analyses in economics, such as measures of dispersion, inequality analysis (comparing means and medians), and summarizing survey data.

Definition: The arithmetic mean (or simply mean) of a set of observations is the sum of all observations divided by the number of observations. It is the most commonly used measure of central tendency and represents the average value of the data.

Computation (brief):

• For ungrouped (raw) data: mean = (sum of observations) / (number of observations).
• For grouped data (discrete or continuous): replace each class by its class mark (midpoint), multiply class marks by their frequencies, sum those products and divide by total frequency.
• Weighted mean: when observations have unequal importance (weights), mean = (sum of weight × observation) / (sum of weights).

Properties and useful results:

• The sum of deviations of observations from their arithmetic mean is zero: Σ(x - x̄) = 0.
• Effect of adding a constant k to every observation: new mean = x̄ + k.
• Effect of multiplying every observation by constant k: new mean = k·x̄.
• Combined mean of two samples: weighted by sample sizes.
• Mean is sensitive to extreme values (outliers); it is not a resistant measure.

Relevance in Economics (Class 11 context): Arithmetic mean is used to find average income, per capita measures, average expenditure, average marks, price indexes (as weighted means), and other summary figures that help describe an economic phenomenon for a group.

Definition: The median is the middle value of an ordered data set that divides the distribution into two equal parts — 50% of observations lie below it and 50% above it. It is a measure of central tendency less affected by extreme values (outliers) than the mean.

When to use: Use the median for raw (individual) data, discrete frequency distributions, and grouped (continuous) frequency distributions. It is especially useful for skewed data (e.g., incomes, house prices) where mean is distorted by outliers.

Raw (unarranged) data: Arrange observations in ascending order and find the middle observation.

• If n is odd, median is the value at position (n+1)/2.
• If n is even, median is the average of values at positions n/2 and (n/2)+1.

Discrete frequency distribution: List distinct values in ascending order with their frequencies, compute cumulative frequencies, and find the value whose cumulative frequency reaches or exceeds n/2 (where n is total frequency). That value is the median.

Grouped (continuous) frequency distribution: For class intervals, identify the median class — the class whose cumulative frequency is ≥ n/2. Use linear interpolation inside that class to estimate the median:

Median = L + [(n/2 − C_f) / f] × h

Where:

• L = lower boundary of the median class (for continuous data use class boundaries, not class limits).
• C_f = cumulative frequency of all classes before the median class.
• f = frequency of the median class.
• h = class width (upper boundary − lower boundary of class).
• n = total number of observations (∑ frequencies).

Properties and remarks:

• Median is unique for a given data set (unless averaging two middle values in even-size raw data).
• It is not affected by the magnitude of extreme values — only their position in order.
• For symmetric distributions mean = median; for skewed distributions they differ (median lies between mode and mean for moderately skewed distributions).
• Median can be located graphically using cumulative frequency curves (ogives) at n/2.

Use in economics (Class 11 relevance): Median income or median wage indicates the midpoint of income distribution and is a better indicator of a typical household than mean when incomes are skewed by very high earners.

Definition: The mode of a data set is the value (or values) that occur most frequently. It is the observation with maximum frequency. Mode is the only measure of central tendency that can be used for qualitative (nominal) data.

Types and properties:

• Ungrouped / discrete data: mode is the observation with the highest frequency. If two values tie for highest frequency the distribution is bimodal; more than two gives multimodal. If all values occur equally, the distribution may be said to have no mode.
• Grouped (continuous) data: exact values are not available; we identify the modal class (class interval with highest frequency) and estimate the mode using a formula.
• Uniqueness: mode may not be unique (can be none, one, or several). It is robust against extreme values (outliers).

When to use mode: For nominal/categorical data (e.g., most preferred color), for skewed distributions where median and mean may be misleading, and for describing the most typical category or value.

How to find mode — summary:

• Ungrouped/discrete data: count frequencies and pick the value with maximum frequency.
• Grouped continuous data: (i) identify the modal class (class with largest frequency), (ii) apply the mode formula for grouped data to estimate the mode within that interval.

Limitations / cautions: Mode gives only the most frequent value and ignores the rest of the distribution. For continuous measurements with many distinct values, mode may be less informative unless data are grouped. The grouped-data mode is an estimate that depends on class width and grouping.

Relation with mean and median (empirical relation): For moderately skewed distributions an approximate relation often used is Mode ≈ 3 × Median − 2 × Mean (this is empirical and not universally accurate).

What are partition values? Partition values divide an ordered data set into equal parts. They are used to describe the relative position of observations and to measure dispersion and skewness. The main partition values are quartiles (4 equal parts), deciles (10 equal parts) and percentiles (100 equal parts).

Quartiles (Q1, Q2, Q3): Q1 is the 25th percentile (first quartile), Q2 is the 50th percentile (median or second quartile), and Q3 is the 75th percentile (third quartile). They split data into four equal parts.

Deciles (D1, D2, ..., D9): The k-th decile Dk is the value below which k/10 of the observations lie (D5 = median).

Percentiles (P1, P2, ..., P99): The p-th percentile Pp is the value below which p% of observations lie. P50 = median, P25 = Q1, P75 = Q3.

Two main methods to find them:

• Ungrouped (individual) data: Sort data in ascending order. Use position formulas to find the location; if the position is fractional, interpolate (take average of neighbors or linear interpolation).
• Grouped (continuous) data: Use cumulative frequencies and linear interpolation inside the class interval that contains the required partition value.

Interpretation and use: Partition values tell you how data are distributed (e.g., 25% of students scored below Q1). Quartiles are common in box-and-whisker plots; percentiles are widely used in test scores, growth charts and income distribution; deciles are used in economic analysis (e.g., income deciles).

Notes: Q2 = D5 = P50. For small data sets some teachers use (n+1) in position formulas; CBSE commonly uses (n+1) method for ungrouped data positions. For grouped data interpolation assumes uniform distribution within the class interval.

Summary: For moderately skewed, unimodal distributions Pearson's empirical relation links the three measures: Mode ≈ 3 × Median − 2 × Mean. This is an approximate relation derived from two measures of skewness and helps identify the direction of skewness and the relative positions of mean, median and mode.

Derivation (brief): Pearson proposed two coefficients of skewness:

• Skewness A: (Mean − Mode) / SD
• Skewness B: 3 × (Mean − Median) / SD

Equating these (assuming the same sign and magnitude for both measures in a moderately skewed distribution) gives: Mean − Mode ≈ 3(Mean − Median) ⇒ Mode ≈ 3×Median − 2×Mean.

Interpretation & conditions:

• If Mean > Median > Mode → distribution is positively (right) skewed.
• If Mean < Median < Mode → distribution is negatively (left) skewed.
• The formula is empirical and approximate. It usually holds for unimodal, moderately skewed distributions. It may fail for symmetric, heavily skewed, or multimodal distributions.

How to use: If you know any two of the three measures for a suitable distribution you can estimate the third using the formula. Use it as a quick check for skew direction rather than an exact identity.

Overview

Measures of central tendency — mean, median and mode — summarise a distribution by a single representative value. Each measure has situations where it is most appropriate. Below is a concise comparison, followed by merits and limitations of each.

Comparison

• Arithmetic Mean: Sum of observations divided by number of observations. It uses all data values, so it is sensitive to every observation.
• Median: Middle value when observations are arranged in order (or average of two middle values if n is even). It is a positional measure and divides data into two equal parts.
• Mode: Most frequently occurring value(s) in the data. It is useful for categorical or discrete data and can be multiple or absent.

When to prefer which

• Use mean when data are roughly symmetric and there are no extreme outliers; good for further algebraic manipulation (e.g., variance).
• Use median when data are skewed or contain outliers (e.g., income, house prices) because median is robust.
• Use mode for qualitative/categorical data (e.g., most common brand) or to identify the most frequent value in discrete distributions.

Merits and Limitations

Arithmetic Mean

• Merits: Easy to compute; uses all observations; algebraically tractable (useful in further statistical formulae); unique.
• Limitations: Affected by extreme values (outliers); not defined for open-ended class intervals without assumption; may be inappropriate for skewed distributions; not meaningful for nominal data.

Median

• Merits: Not affected by extreme values; always exists and is unique for a finite list of numbers; applicable to ordinal, interval and ratio data; appropriate for skewed distributions.
• Limitations: Does not use all information in the data (ignores magnitude of values); not amenable to algebraic manipulation as easily as mean; for grouped data median is estimated (not exact).

Mode

• Merits: Simple to understand; applicable to nominal (categorical) data; shows the most typical category or value; can exist with frequencies in raw data easily.
• Limitations: May not be unique (bimodal, multimodal) or may not exist; ignores other values and magnitudes; unstable for small changes in data; for grouped data mode is an estimate.

Practical considerations

• For symmetric, well-behaved numeric data: mean is preferred.
• For skewed distributions or incomes/wealth: median better represents a 'typical' case.
• For categorical data or most-common-value questions: mode is appropriate.
• Always state which measure you are using and why — different measures can lead to different policy or business conclusions (e.g., mean salary versus median salary).

What this topic covers
Practical considerations and applications explain how to choose and use mean, median and mode correctly in real data analysis (especially economic data), their limitations, and the interpretation of results for policy and planning.

How to choose a measure

• Scale of measurement: use mode for nominal data, median for ordinal data, and mean for interval/ratio data when values are symmetric and not severely skewed.
• Effect of outliers/skewness: the mean is sensitive to extreme values; median is robust and better represents a typical value when distribution is skewed (e.g., income).
• Purpose of analysis: use mode to find the most common category (e.g., common shoe size), median to find the central position (e.g., middle income), mean for overall average (e.g., average production per factory).

Practical accounting for data form

• Ungrouped (raw) data: direct formulas for mean, median, mode apply.
• Grouped (frequency distribution): use class mid‑points for mean, ogives to estimate median, and identify modal class for mode — use grouped formulas for more accurate estimates.
• Open‑ended intervals: mean cannot be computed exactly if both tails are open; median can often be estimated if cumulative frequencies are known.

Transformations and properties useful in practice

• If each observation x_i is changed to x_i + a, mean and median increase by a; mode shifts by a. If multiplied by k, they all multiply by k (k>0).
• The weighted mean is used when observations have different importance (weights), e.g., computing CPI or index numbers.

Limitations and checks

• Mean does not describe distribution shape — accompany mean with dispersion measures (range, variance, quartiles) and graphical summary.
• Mode may be non-unique or unstable in continuous data; choose modal class for grouped data.
• Always consider sample size, measurement scale, and presence of errors or outliers when reporting central tendency.

Applications in economics and daily life
Central tendency measures are used in reporting average income, median wage, most common household size, average marks, consumer price indices, production averages, and for policy decisions such as setting minimum wage, subsidy targeting, and taxation thresholds. Combine with graphs (histogram, box-plot) and dispersion measures for sound conclusions.