Different averages answer different questions. The correct average depends on whether we want a general value, a middle position, the most common item, a growth rate or a rate such as speed.
Meaning and Need for an Average
Suppose the marks of 100 students are available. Reading all 100 values does not immediately tell us the general performance of the class. A single representative value helps us understand the overall level of the data.
Why Do We Need an Average?
- It summarises a large mass of data.
- It helps compare two or more groups.
- It gives a quick idea of the general level of observations.
- It supports business, economic and statistical decisions.
From Data to Decision
Qualities of an Ideal Measure
Clearly Defined
Different persons should obtain the same answer from the same data.
Simple
It should be easy to understand and reasonably easy to calculate.
Uses the Data
It should use all observations wherever the nature of the measure permits.
Reliable
It should be stable, useful for comparison and suitable for further analysis.
Which Average Should You Use?
Arithmetic Mean
Use when all numerical observations should contribute to one general average.
Memory rule: Ordinary averageMedian
Use when position matters or extreme values and open-ended classes are present.
Memory rule: Middle positionMode
Use when the most common size, choice, brand or value is required.
Memory rule: Most frequentUse Geometric Mean When
- Growth takes place over several periods.
- Rates combine multiplicatively.
- Compound returns or index numbers are involved.
Use Harmonic Mean When
- Average speed is required over equal distances.
- Rates or “per-unit” quantities are averaged.
- The variable appears in the denominator.
| Situation | Best Measure | Reason |
|---|---|---|
| General numerical data | Arithmetic Mean | Uses all observations and permits algebraic treatment. |
| Income data with a few very rich persons | Median | Extreme values do not pull the result heavily. |
| Most demanded shoe size | Mode | Identifies the value occurring most frequently. |
| Average annual compound growth | Geometric Mean | Correct for multiplicative growth. |
| Average speed for equal distances | Harmonic Mean | Correct for rates with equal quantities in the numerator. |
Property Comparison: What Each Average Can and Cannot Do
ICAI frequently asks these properties directly. Learn the table as a decision sheet rather than trying to derive the answer during the examination.
| Property | AM | Median | Mode | GM | HM |
|---|---|---|---|---|---|
| Rigidly or uniquely defined | Yes | Yes | Not always | Yes | Yes |
| Based on all observations | Yes | No | No | Yes | Yes |
| Affected by extreme values | Yes | No | Generally little | Yes | Yes, especially by small values |
| Suitable for open-ended classes | No | Yes | Often possible | No | No |
| Capable of algebraic treatment | Excellent | Limited | Limited | Yes | Yes |
| Can be located graphically | No | Yes, by ogive | Yes, by histogram | No | No |
| Main use | General average | Position and skewed data | Most common value | Growth and ratios | Rates and equal-distance speed |
Arithmetic Mean
Arithmetic Mean is obtained by adding all observations and dividing by the number of observations.
x̄ = Σx / n
| Symbol | Meaning |
|---|---|
| x̄ | Arithmetic Mean of the observations. |
| Σ | “Sum of”. It tells us to add all the values that follow it. |
| x | Each individual observation or value. |
| Σx | Total of all individual observations. |
| n | Number of individual observations. |
x̄ = Σfx / Σf = Σfx / N
| Symbol | Meaning |
|---|---|
| x̄ | Arithmetic Mean of the distribution. |
| f | Frequency corresponding to a value or class. |
| x | Value in a discrete series or midpoint of a class in grouped data. |
| fx | Product of frequency and the corresponding value or class midpoint. |
| Σfx | Total of all the fx products. |
| Σf or N | Total frequency, meaning the total number of observations. |
Individual Data
Fully Solved Illustration: Arithmetic Mean of Individual Data
Question: Daily wages of nine workers are ₹58, ₹62, ₹48, ₹53, ₹70, ₹52, ₹60, ₹84 and ₹75. Find the arithmetic mean wage.
Interpretation: The nine workers earn an average daily wage of approximately ₹62.44.
Grouped Frequency Distribution
For grouped data, first calculate the midpoint of every class. Multiply the midpoint by the corresponding frequency and then divide the total of fx by the total frequency.
| Weight (kg) | Frequency (f) | Midpoint (x) | fx |
|---|---|---|---|
| 44–48 | 3 | 46 | 138 |
| 49–53 | 4 | 51 | 204 |
| 54–58 | 5 | 56 | 280 |
| 59–63 | 7 | 61 | 427 |
| 64–68 | 9 | 66 | 594 |
| 69–73 | 8 | 71 | 568 |
| Total | 36 | — | 2211 |
Fully Solved Illustration: Arithmetic Mean of Grouped Data
Interpretation: The estimated average weight of the 36 persons is 61.42 kg.
Step-Deviation Method
| Symbol | Meaning |
|---|---|
| x̄ | Arithmetic Mean to be calculated. |
| A | Assumed Mean, normally a convenient central value or midpoint. |
| x | Value or class midpoint. |
| C | Common factor or common class width used to simplify deviations. |
| d | Step deviation, calculated as (x − A) / C. |
| f | Frequency corresponding to each value or class. |
| fd | Product of frequency and step deviation. |
| Σfd | Total of all fd values. |
| N | Total frequency, Σf. |
Fully Solved Illustration: Step-Deviation Method
Question: Find the mean for class midpoints 10, 20, 30, 40 and 50 with frequencies 2, 3, 5, 4 and 1.
| x | f | d = (x − 30) ÷ 10 | fd |
|---|---|---|---|
| 10 | 2 | −2 | −4 |
| 20 | 3 | −1 | −3 |
| 30 | 5 | 0 | 0 |
| 40 | 4 | 1 | 4 |
| 50 | 1 | 2 | 2 |
| Total | 15 | — | −1 |
Properties of Arithmetic Mean and Combined Mean
| Property | Meaning |
|---|---|
| Constant observations | If every observation equals k, the mean is also k. |
| Sum of deviations | Σ(x − x̄) = 0 and Σf(x − x̄) = 0. |
| Change of origin and scale | If y = a + bx, then ȳ = a + bx̄. |
| Combined mean | Means of two or more groups are combined using their group sizes. |
x̄ = (n₁x̄₁ + n₂x̄₂) / (n₁ + n₂)
| Symbol | Meaning |
|---|---|
| x̄ | Combined Mean of both groups. |
| n₁ and n₂ | Number of observations in the first and second groups. |
| x̄₁ and x̄₂ | Mean of the first and second groups. |
| n₁x̄₁ and n₂x̄₂ | Total value represented by each group. |
| n₁ + n₂ | Total number of observations in both groups together. |
Combined Mean Example
40 female workers earn an average of ₹5,200 and 60 male workers earn an average of ₹6,800.
Change of Origin and Scale: Direct Transformation Questions
ICAI often gives a relationship such as y = a + bx and asks for the Mean, Median or Mode of y. Do not recalculate the complete series.
Mean(y) = a + b Mean(x)
Median(y) = a + b Median(x)
Mode(y) = a + b Mode(x)
| Symbol | Meaning |
|---|---|
| x | Original variable. |
| y | Transformed variable. |
| a | Constant added to every observation; change of origin. |
| b | Constant multiplying every observation; change of scale. |
Fully Solved Illustration: Median under Transformation
Question: y = 2x − 3 and Median(x) = 20. Find Median(y).
Fully Solved Illustration: Mean from an Equation
Question: 2u + v + 7 = 0 and Mean(u) = 10. Find Mean(v).
Missing Frequency Problems
| Symbol | Meaning |
|---|---|
| Σf | Total of all known and unknown frequencies. |
| N | Total number of observations. |
| Σfx | Total of frequency multiplied by the corresponding value. |
| x̄ | Given Arithmetic Mean. |
| Unknown frequency | The missing value is usually represented by a letter such as m, x or y. |
Fully Solved Illustration: Missing Frequency
Question: Values 10, 20, 30 and 40 have frequencies 3, 5, m and 2. If the mean is 25, find m.
| x | f | fx |
|---|---|---|
| 10 | 3 | 30 |
| 20 | 5 | 100 |
| 30 | m | 30m |
| 40 | 2 | 80 |
| Total | 10 + m | 210 + 30m |
Median
Median is the middle-most value after arranging observations in ascending or descending order. It is a positional average.
Odd Number of Observations
Median position = (n + 1) / 2 th item.
Example: For 7 observations, median is the 4th item.
Even Number of Observations
Median is the average of the n/2 th and (n/2 + 1) th items.
Example: For 8 observations, average the 4th and 5th items.
| Symbol | Meaning in Individual-Series Median |
|---|---|
| n | Total number of observations after arranging them in order. |
| (n + 1) / 2 | Position of the median when the number of observations is odd. |
| n / 2 and (n / 2) + 1 | The two middle positions when the number of observations is even. |
| Median | For even observations, the average of the values at the two middle positions. |
Odd Number Example
Marks: 72, 85, 56, 80, 65, 52, 68
Ordered data: 52, 56, 65, 68, 72, 80, 85
Even Number Example
Ordered wages: 56, 82, 82, 96, 100, 106, 110, 120
Grouped Frequency Distribution
| Symbol | Meaning |
|---|---|
| L | Lower class boundary of the median class. |
| N | Total frequency, Σf. |
| N/2 | Position of the median observation. |
| c.f. | Cumulative frequency of the class immediately preceding the median class. |
| f | Frequency of the median class. |
| C | Class width or class size of the median class. |
Fully Solved Illustration: Median of Grouped Data
Question: Find the median of the following distribution.
| Class | f | Cumulative f |
|---|---|---|
| 0–10 | 5 | 5 |
| 10–20 | 9 | 14 |
| 20–30 | 12 | 26 |
| 30–40 | 8 | 34 |
| 40–50 | 6 | 40 |
Interpretation: Half the observations lie below approximately 25 and half lie above it.
Properties of Median
- If y = a + bx, then Median of y = a + b × Median of x.
- The sum of absolute deviations Σ|x − A| is minimum when A is the median.
- Median is suitable for open-ended and highly skewed distributions.
Partition Values
Partition values divide an ordered distribution into equal parts.
Quartiles
Divide the distribution into four equal parts.
Q₁, Q₂ and Q₃Deciles
Divide the distribution into ten equal parts.
D₁ to D₉Percentiles
Divide the distribution into one hundred equal parts.
P₁ to P₉₉Unclassified Data
| Symbol | Meaning |
|---|---|
| n | Total number of ordered observations. |
| p | Required proportion of the distribution. |
| k | Number of the required quartile, decile or percentile. |
| p = k/4 | Used for quartiles, such as Q₁ where p = 1/4. |
| p = k/10 | Used for deciles, such as D₇ where p = 7/10. |
| p = k/100 | Used for percentiles, such as P₆₀ where p = 60/100. |
Use p = 1/4, 2/4 or 3/4 for quartiles; p = k/10 for deciles; and p = k/100 for percentiles.
Fully Solved Illustration: Quartile in Individual Data
Question: Find Q₁ for the ordered values 5, 8, 12, 15, 18, 21, 25, 30, 34 and 40.
Graphical Determination of Median and Quartiles
Median, quartiles, deciles and percentiles are positional values and can be located from an ogive, that is, a cumulative-frequency curve.
Grouped Data
| Symbol | Meaning |
|---|---|
| L | Lower class boundary of the class containing the required partition value. |
| N | Total frequency, Σf. |
| p | Required proportion, such as 1/4 for Q₁, 7/10 for D₇ or 60/100 for P₆₀. |
| Np | Position of the required quartile, decile or percentile. |
| c.f. | Cumulative frequency of the class immediately preceding the relevant class. |
| f | Frequency of the class containing the partition value. |
| C | Class width or class size. |
Fully Solved Illustration: First Quartile of Grouped Data
Using the grouped distribution in the median example, find Q₁.
Mode
Mode is the value that occurs most frequently. It represents the most common or most popular value.
Mode is Useful For
- Most demanded shoe or garment size.
- Most popular brand or model.
- Most common wage, price or choice.
Possible Forms
- Unimodal: one mode.
- Bimodal: two modes.
- Multimodal: more than two modes.
- No mode: equal frequencies.
Simple Example
5, 3, 8, 9, 5, 6
Grouped Frequency Distribution
| Symbol | Meaning |
|---|---|
| L | Lower class boundary of the modal class. |
| f₀ | Frequency of the modal class, which is normally the highest frequency. |
| f₋₁ | Frequency of the class immediately preceding the modal class. |
| f₁ | Frequency of the class immediately succeeding the modal class. |
| C | Class width or class size of the modal class. |
The modal class is the class with the highest frequency. The formula estimates the position of the mode within that class.
Fully Solved Illustration: Mode of Grouped Data
Question: Frequencies for classes 0–10, 10–20, 20–30, 30–40 and 40–50 are 4, 7, 12, 9 and 3. Find the mode.
Empirical Relationship
Mode = 3 Median − 2 Mean
| Term | Meaning |
|---|---|
| Mean | Arithmetic Mean of the distribution. |
| Median | Middle positional value of the distribution. |
| Mode | Most frequently occurring or most typical value. |
| 3 | Empirical coefficient used in the approximate relationship for a moderately skewed distribution. |
Missing Frequencies Using Median or Mode
When frequencies are missing, first use the total-frequency condition. Then use the given Median or Mode formula to form another equation.
Illustration: Missing Frequencies when Median is Given
A distribution has total frequency 100. Two frequencies are missing. The Median is 32.
Exam point: Do not guess the missing frequencies. The two equations must be solved simultaneously.
Illustration: Missing Frequency when Mode is Given
If the modal value is given, identify the modal class first. The unknown frequency may appear as f₀, f₋₁ or f₁.
Geometric Mean
For n positive observations, Geometric Mean is the nth root of their product.
| Symbol | Meaning |
|---|---|
| G | Geometric Mean. |
| x₁, x₂, …, xₙ | Positive individual observations. |
| n | Number of observations. |
| 1/n | Indicates that the nth root of the product is to be taken. |
| Symbol | Meaning |
|---|---|
| G | Geometric Mean. |
| log G | Logarithm of the Geometric Mean. |
| x | Positive value or class midpoint. |
| log x | Logarithm of each value or class midpoint. |
| f | Frequency corresponding to each value. |
| Σf log x | Total of frequency multiplied by the logarithm of the corresponding value. |
| N | Total frequency, Σf. |
| Antilog | Reverse logarithm used to obtain G after calculating log G. |
Fully Solved Illustration: Geometric Mean of Individual Data
Question: Find the GM of 3, 6 and 12.
Fully Solved Illustration: Average Compound Growth Rate
Sales grow by 10%, 20% and 5% in three successive years. Find the average annual compound growth rate.
The compound average is approximately 11.5% per year, not the simple average of 11.67%.
Use GM For
- Compound growth rates.
- Average investment returns.
- Index numbers and ratios.
Remember
- All observations should be positive.
- GM is suitable for multiplicative change.
- Logarithms simplify its calculation.
Harmonic Mean
Harmonic Mean is the reciprocal of the arithmetic mean of the reciprocals.
| Symbol | Meaning |
|---|---|
| H | Harmonic Mean. |
| n | Number of individual observations. |
| x | Each positive observation. |
| 1/x | Reciprocal of each observation. |
| Σ(1/x) | Total of the reciprocals of all observations. |
| Symbol | Meaning |
|---|---|
| H | Harmonic Mean. |
| N | Total frequency, Σf. |
| f | Frequency corresponding to each value. |
| x | Value or class midpoint. |
| f/x | Frequency divided by the corresponding value. |
| Σ(f/x) | Total of all f/x values. |
Fully Solved Illustration: Harmonic Mean of Individual Data
Question: Find the HM of 4, 6 and 10.
Fully Solved Illustration: Average Speed over Equal Distances
A vehicle covers 120 km at 40 km/h and another 120 km at 60 km/h.
This is also obtained by HM: 2 ÷ (1/40 + 1/60) = 48 km/h.
Equal Distances
Use Harmonic Mean for average speed.
Example: Equal kilometres at 40 km/h and 60 km/h.
Equal Times
Use Arithmetic Mean for average speed.
Example: One hour at 40 km/h and one hour at 60 km/h.
Relationship between AM, GM and HM
| Symbol | Meaning |
|---|---|
| AM | Arithmetic Mean. |
| GM | Geometric Mean. |
| HM | Harmonic Mean. |
| ≥ | Greater than or equal to. |
For the same set of positive observations, AM cannot be smaller than GM, and GM cannot be smaller than HM. Equality occurs only when all observations are equal.
Fully Solved Illustration: Verify AM ≥ GM ≥ HM
For the observations 6, 8, 12 and 36:
Therefore, 15.5 ≥ 12 ≥ 9.93.
Special Result for Two Positive Numbers
| Symbol | Meaning |
|---|---|
| AM | Arithmetic Mean of exactly two positive observations. |
| GM | Geometric Mean of the same two observations. |
| HM | Harmonic Mean of the same two observations. |
| GM² | Square of the Geometric Mean. |
Weighted Averages
When observations do not have equal importance, a weight is attached to each observation.
| Symbol | Meaning |
|---|---|
| x | Observation or value. |
| w | Weight or relative importance attached to the observation. |
| wx | Product of the weight and the observation. |
| Σwx | Total of all weighted values. |
| Σw | Total of all weights. |
| Symbol | Meaning |
|---|---|
| x | Positive observation or growth relative. |
| w | Weight attached to each observation. |
| log x | Logarithm of each observation. |
| Σw log x | Total of weight multiplied by the logarithm of each observation. |
| Σw | Total of all weights. |
| Antilog | Reverse logarithm used to convert the result back to the original scale. |
| Symbol | Meaning |
|---|---|
| x | Positive observation, rate or price. |
| w | Weight attached to each observation. |
| w/x | Weight divided by the corresponding observation. |
| Σ(w/x) | Total of all w/x values. |
| Σw | Total of all weights. |
Fully Solved Illustration: Weighted Arithmetic Mean
A student scores 70, 80 and 90 in components carrying weights of 20%, 30% and 50%.
| Score x | Weight w | wx |
|---|---|---|
| 70 | 20 | 1,400 |
| 80 | 30 | 2,400 |
| 90 | 50 | 4,500 |
| Total | 100 | 8,300 |
Fully Solved Illustration: Weighted Harmonic Mean
A buyer purchases 20 units at ₹10 per unit and 30 units at ₹15 per unit. Find the weighted harmonic mean price where quantities are the weights.
Quick Comparative Review
| Measure | Main Strength | Main Weakness | Best Use |
|---|---|---|---|
| AM | Uses all observations and is algebraically useful | Affected by extreme values | General numerical analysis |
| Median | Resists extremes and works with open ends | Does not use every value fully | Income and skewed data |
| Mode | Shows the most common value | May be absent or multiple | Popular size, brand or choice |
| GM | Correct for compound growth | Cannot ordinarily use zero or negative values | Growth rates and index numbers |
| HM | Correct for rates and ratios | Highly affected by very small values | Speed and per-unit rates |
Worked Examples and Exam Traps
Example 1: Mode from Mean and Median
Mean = 55.60 and Median = 52.40.
Example 2: Two Numbers from AM and GM
For two positive numbers, AM = 5 and GM = 4.
Example 3: Average Speed
A vehicle covers equal distances at 40 km/h and 60 km/h.
Question-Type Coverage Checklist
The following table converts the exercise pattern into direct revision instructions. A student should be able to answer each line without deriving the rule during the examination.
| Question Type | Direct Rule to Recall |
|---|---|
| Meaning of central tendency | It measures the central location, not dispersion or scatterness. |
| Grouped AM assumption | All observations in a class are represented by the class midpoint. |
| Sampling fluctuation | AM is affected by sampling fluctuations; the statement “not affected” is wrong. |
| Open-end classification | Median is generally the best measure. |
| Extreme observations | Median is not materially affected; AM is affected. |
| Even number of observations | Median is the simple average of the two middle ordered values. |
| Uniqueness | Mode may not be uniquely defined. |
| Average rates | GM and HM are considered; HM is especially relevant for rates with equal quantities. |
| Profits and losses | GM cannot ordinarily be used when values include negatives. |
| Graphical quartiles | Use an ogive. |
| Linear transformation | Apply y = a + bx directly to Mean, Median and Mode. |
| AM = GM | All positive observations are equal, therefore HM is also equal. |
Solved Numerical Drill
1. Find Two Numbers from AM and GM
AM = 6.5 and GM = 6 for two positive numbers.
2. Combined Mean and Group Proportion
Unskilled workers earn ₹10,000 on average, skilled workers earn ₹15,000, and the combined mean is ₹12,000. Find the percentage of skilled workers.
3. Combined Harmonic Mean
Two groups contain 15 and 13 observations and have HMs of 75 and 65 respectively.
4. Equal-Distance Average Speed
An aircraft travels from A to B at 500 km/h and returns over the same distance at 700 km/h.
Common Exam Traps
| Trap | Correct Idea |
|---|---|
| Mean is unaffected by extremes | Wrong. Mean is strongly affected by extreme values. |
| Median class has the highest frequency | Wrong. It contains the N/2th item. |
| Mode always exists | Wrong. It may be absent or more than one. |
| GM can ordinarily use negative values | Wrong in the usual CA Foundation treatment. |
| Average speed is always AM | Use HM when equal distances are covered. |
| AM × HM = GM² for every data set | It holds only for two positive observations. |
Measures of Central Tendency — One Page Recall
- One representative value summarises the data.
- AM for general numerical use.
- Median for position, open ends and skewed data.
- Mode for the most common value.
- GM for growth and HM for rates.
- x̄ = Σx/n or Σfx/N.
- Uses every observation.
- Σ(x − x̄) = 0.
- Suitable for algebraic treatment.
- Affected by extreme values.
- Middle positional value.
- Arrange data before locating it.
- Grouped data uses N/2.
- Works with open-ended classes.
- Minimises absolute deviations.
- Quartiles divide into 4 parts.
- Deciles divide into 10 parts.
- Percentiles divide into 100 parts.
- Q₂ = D₅ = P₅₀ = Median.
- Most frequently occurring value.
- Modal class has the highest frequency.
- Mode ≈ 3 Median − 2 Mean.
- May be absent or multiple.
- Used for compound and multiplicative change.
- Useful for growth rates and index numbers.
- Requires positive observations.
- Logarithms simplify calculation.
- Reciprocal of the AM of reciprocals.
- Used for rates and per-unit quantities.
- Equal distances → HM for average speed.
- Equal times → AM for average speed.
- For positive data: AM ≥ GM ≥ HM.
- Equality when all observations are equal.
- For two positive values: AM × HM = GM².
- Weighted AM = Σwx/Σw.
- Larger weight means greater influence.