Unit 1: Descriptive statistics
Quantitative Techniques notes · PTU syllabus (MBA 103-18)
On this page
Unit summary
Descriptive statistics summarise data so that patterns become visible. This unit covers the meaning, features, functions, scope and limitations of statistics, measures of central tendency and partition values, measures of variation and their coefficients, moments, skewness (Karl Pearson and Bowley) and kurtosis.
After this unit you can
- Explain the meaning, functions, scope and limitations of statistics
- Compute measures of central tendency and partition values
- Compute absolute and relative measures of dispersion
- Compute moments, skewness and kurtosis and interpret them
PTU syllabus topics
- Meaning
- features
- functions and limitations of statistics
- measures of central tendency (arithmetic/geometric/harmonic mean, mode, median, quartiles/deciles/percentiles)
- measures of variation (range, quartile deviation, mean deviation, standard deviation)
- moments
- skewness and kurtosis
Topic 1
Meaning, features, importance, functions and scope
Statistics (singular) is the science of collecting, classifying, presenting, analysing and interpreting numerical data. Statistics (plural) are numerical facts collected for a purpose. Features (Secrist): aggregates of facts, numerically expressed, affected by multiple causes, collected systematically for a purpose, reasonably accurate and comparable. Importance in business: market research, demand forecasting, quality control, financial analysis, HR analytics (attrition, performance) and decision-making under uncertainty. Limitations: studies only quantitative data and groups (not individuals), results are true on average, and data can be misused.
- 1
Presents facts in definite form
- 2
Simplifies mass data
- 3
Facilitates comparison
- 4
Helps in formulating and testing hypotheses
- 5
Helps in prediction and forecasting
- 6
Helps in policy formulation
Example
Statistics in business: a retail chain collects daily sales from 50 stores, classifies them by region, presents them in charts, finds averages and trends, and uses forecasts to plan stock — every function of statistics in one use.
Exam tip
Functions: simplifying data, comparison, forecasting, testing hypotheses and policy-making. Limitations: deals with groups, not individuals, and can be misused.
Topic 2
Measures of central tendency
The long tail on the left pulls the mean down, so Mean < Median < Mode
The long tail on the right pulls the mean towards the tail, so Mean > Median > Mode
In a symmetric bell-shaped curve all three averages coincide at the centre
Arithmetic mean
x̄ = Σfx / Σf
Geometric mean
GM = antilog (Σ log x / n)
For growth rates and ratios
Harmonic mean
HM = n / Σ(1/x)
For rates and speeds
Median (grouped)
L + [(N/2 − cf) / f] × h
Mode (grouped)
L + [(f₁ − f₀) / (2f₁ − f₀ − f₂)] × h
- Relationship: for positive values, AM ≥ GM ≥ HM; for moderately skewed data, Mode = 3 Median − 2 Mean.
- Choosing an average: mean for symmetric data; median for skewed data or open-ended classes (incomes); mode for the most common size or brand.
Example
Mean, median and mode of 4, 6, 6, 8, 16: mean = 40 ÷ 5 = 8. The median is the middle value, 6. The mode is the most frequent value, 6. The mean is pulled up by the extreme value 16, so the median represents this data better.
Exam tip
Use the median for skewed data such as income, and the mode for the most popular size or brand.
Topic 3
Partition values
Partition values divide ordered data into equal parts:
- Quartiles (Q1, Q2, Q3) divide into 4 parts; Q2 = median. Qk = L + [(kN/4 − cf) / f] × h.
- Deciles (D1 … D9) divide into 10 parts; percentiles (P1 … P99) into 100.
Example
If a student's score is at the 90th percentile, 90% of students scored below that student.
Topic 4
Measures of dispersion
Both curves are centred on the same average; the narrow tall curve has small variation, the wide flat curve large variation
| Measure | Formula | Relative measure |
|---|---|---|
| Range | L − S | (L − S)/(L + S) |
| Quartile deviation | (Q3 − Q1)/2 | (Q3 − Q1)/(Q3 + Q1) |
| Mean deviation | Σf abs(x − A)/N (A = mean or median) | MD / A |
| Standard deviation | √[Σf(x − x̄)²/N] | CV = σ/x̄ × 100 |
| Variance | σ² | — |
Example
Two shops have average daily sales of ₹10,000. Shop A's SD is ₹500 and Shop B's is ₹2,500, so Shop A's sales are far more stable.
Exam tip
Standard deviation is the best measure of dispersion because it uses all values and allows further algebraic treatment.
Topic 5
Moments
Moments are averages of the powers of deviations — they describe the shape of a distribution.
First central moment
μ1 = Σf(x − x̄) ÷ N = 0
Second central moment
μ2 = Σf(x − x̄)² ÷ N = variance
Third central moment
μ3 = Σf(x − x̄)³ ÷ N — skewness
Fourth central moment
μ4 = Σf(x − x̄)⁴ ÷ N — kurtosis
Conversion from raw moments
μ2 = μ2′ − μ1′²; μ3 = μ3′ − 3μ2′μ1′ + 2μ1′³; μ4 = μ4′ − 4μ3′μ1′ + 6μ2′μ1′² − 3μ1′⁴
Example
Central moments of 2, 4, 6: the mean is 4 and the deviations are −2, 0 and 2. μ₁ = 0. μ₂ = (4 + 0 + 4) ÷ 3 = 2.67, the variance. μ₃ = (−8 + 0 + 8) ÷ 3 = 0, so the distribution is symmetrical.
Exam tip
μ₁ is always zero, μ₂ is the variance, μ₃ measures skewness and μ₄ measures kurtosis.
Topic 6
Skewness
Skewness measures the lack of symmetry. In a positively skewed distribution, Mean > Median > Mode (long right tail); in a negatively skewed one, Mean < Median < Mode.
Karl Pearson's coefficient
Sk = (Mean − Mode) ÷ σ, or 3 (Mean − Median) ÷ σ
Bowley's coefficient (quartiles)
Sk = (Q3 + Q1 − 2Median) ÷ (Q3 − Q1)
Moment coefficient
β1 = μ3² ÷ μ2³; γ1 = √β1 with the sign of μ3
Example
Mean 52, median 50, σ 8: Karl Pearson's Sk = 3 × (52 − 50) ÷ 8 = 0.75 — moderately positively skewed. Q1 = 40, Q3 = 66, median 50: Bowley's Sk = (66 + 40 − 100) ÷ 26 = 0.23.
- Karl Pearson's measure lies (practically) between −3 and +3; Bowley's between −1 and +1.
Topic 7
Kurtosis
Kurtosis measures the peakedness (or tail-heaviness) of a distribution relative to the normal curve.
Coefficient
β2 = μ4 ÷ μ2²
Excess kurtosis
γ2 = β2 − 3
Mesokurtic
β2 = 3 — normal curve
Leptokurtic
β2 > 3 — more peaked, heavy tails (stock returns)
Platykurtic
β2 < 3 — flatter than normal
Exam tip
Skewness tells the direction of asymmetry; kurtosis tells the peakedness — a four-measure summary (mean, SD, skewness, kurtosis) describes any distribution well.
Example
Reading β₂: if μ₂ = 4 and μ₄ = 60, then β₂ = μ₄ ÷ μ₂² = 60 ÷ 16 = 3.75. Since β₂ is greater than 3, the distribution is leptokurtic, more peaked than the normal curve.
Exam tip
β₂ = 3 mesokurtic (normal), > 3 leptokurtic, < 3 platykurtic.
Key terms
- Central tendency
- A single value representing the centre of data
- Coefficient of variation
- SD as a percentage of the mean
- Moment
- Average of powered deviations from the mean
- Skewness
- Lack of symmetry in a distribution
- Kurtosis
- Degree of peakedness of a distribution
Quick revision
- Averages: AM, GM, HM, median, mode; partition values Q, D, P.
- Dispersion: range, QD, MD, SD; CV for comparison.
- Moments μ1 to μ4; μ2 = variance.
- Skewness: Karl Pearson, Bowley, moments.
- Kurtosis: β2 = 3 meso, > 3 lepto, < 3 platy.
Important exam questions
Practice questions written to the PTU exam pattern for this unit's syllabus: short answers (Section A style) and long answers (Sections B and C style).
Short-answer questions
- Q1.State four limitations of statistics.
- Q2.What are partition values?
- Q3.Define coefficient of variation.
- Q4.What are central moments?
- Q5.Distinguish Karl Pearson's and Bowley's coefficients of skewness.
- Q6.What is a leptokurtic curve?
Long-answer questions
- Q1.Explain the meaning, functions, scope and limitations of statistics.
- Q2.Explain measures of central tendency and their merits.
- Q3.Explain measures of dispersion with a numerical example.
- Q4.Explain moments, skewness and kurtosis with their measures.
Stuck on this unit?
Message SBS on WhatsApp for help with Quantitative Techniques, or to ask about studying MBA at Synetic.
