Comprehensive Descriptive Statistics Processor
Compute complete population and sample descriptive statistics: Mean (Arithmetic, Geometric, Harmonic), Median, Mode, Variance, Standard Deviation, Interquartile Range (IQR), Skewness, Kurtosis, and Tukey 5-number summaries.
# Calculation Pipeline (Chain of Thought)
- Step 1: Dataset Sanitization & Monotonic Sort: Ingest numerical vector
X = [x_1, x_2, dots, x_n] of length n, sort in ascending order x_{(1)} le x_{(2)} le dots le x_{(n)}, and compute total sum sum x_i. - Step 2: Central Tendency Evaluation: Arithmetic Mean:
ar{x} = rac{1}{n}sum_{i=1}^n x_i; Median: M = x_{( rac{n+1}{2})} for odd n, or rac{x_{(n/2)} + x_{(n/2 + 1)}}{2} for even n; Mode: ext{argmax}_{v}( ext{count}(v)). - Step 3: Dispersion & Bessel\'s Sample Variance: Sum of Squared Deviations:
SS = sum_{i=1}^n (x_i - ar{x})^2. Sample Variance: s^2 = rac{SS}{n - 1}; Sample Std Dev: s = sqrt{s^2}. Population Std Dev: sigma = sqrt{SS / N}. - Step 4: Tukey Quartiles, Skewness & Outliers: Compute
Q_1 = P_{25}, Q_3 = P_{75}, ext{IQR} = Q_3 - Q_1. Classify Tukey outliers outside [Q_1 - 1.5cdot ext{IQR},; Q_3 + 1.5cdot ext{IQR}] and evaluate Fisher-Pearson Skewness g_1 = rac{m_3}{s^3}.
Decision Rule: Central Tendency Rule: For symmetric Gaussian distributions, report the Mean and Standard Deviation. For skewed datasets or distributions containing extreme outliers, report the Median and Interquartile Range (IQR) for robust reporting.
+-------------------------------------------------------------------------+
| DESCRIPTIVE STATISTICS PIPELINE |
+-------------------------------------------------------------------------+
| Raw Data: [12, 7, 22, 15, 18, 9, 31, 14, 15] (n = 9) |
| Sorted: [7, 9, 12, 14, 15, 15, 18, 22, 31] |
| |
| Central Tendency: |
| - Sum (Σx): 143 |
| - Mean (x̄): 143 / 9 = 15.89 |
| - Median (Q2): 15.00 (5th element of 9) |
| - Mode: 15.00 (frequency = 2) |
| |
| Dispersion & Spread: |
| - Sum of Squares (SS) = Σ(xi - 15.89)^2 = 422.89 |
| - Sample Var (s^2) = 422.89 / (9 - 1) = 52.86 |
| - Sample Std Dev (s) = sqrt(52.86) = 7.27 |
| - Population Std (σ) = sqrt(422.89 / 9) = 6.85 |
| |
| Tukey 5-Number Summary: |
| - Min: 7 | Q1: 10.5 | Median: 15 | Q3: 20.0 | Max: 31 |
| - IQR = 20.0 - 10.5 = 9.5 |
| - Fences: [10.5 - 1.5(9.5), 20.0 + 1.5(9.5)] = [-3.75, 34.25] |
| - Outliers: None (31 falls within upper fence 34.25) |
+-------------------------------------------------------------------------+
Frequently Asked Questions
When should I use the Median instead of the Mean to describe data?
The Mean is sensitive to extreme outliers and asymmetric skews (e.g., real estate prices, net worth distributions, household income). A single billionaire skews the mean drastically upward, whereas the Median represents the true 50th percentile midpoint unaffected by extreme values. Use Mean for symmetric normal distributions and Median for skewed data.
Why does Bessel's correction divide by (n - 1) instead of n for sample variance?
When calculating variance from an empirical sample, deviations are measured against the sample mean (x̄) rather than the true unknown population mean (μ). The sample mean inherently minimizes the sum of squared deviations for that specific sample, creating an underestimate (bias). Dividing by (n - 1) corrects this degrees-of-freedom bias, making s^2 an unbiased estimator of σ^2.
How does the Interquartile Range (IQR) identify statistical outliers?
The IQR represents the spread of the middle 50% of the data (Q3 - Q1). Under John Tukey's classic exploratory data analysis method, inner fences are set at Q1 - 1.5*IQR (lower) and Q3 + 1.5*IQR (upper). Any observation falling outside these boundaries is flagged as an outlier, and beyond 3.0*IQR as an extreme outlier.
What do Skewness and Kurtosis tell you about a dataset's distribution?
Skewness measures distribution asymmetry: a positive skew (>0) indicates a long tail to the right with values concentrated on the left, while a negative skew (<0) indicates a long tail to the left. Kurtosis measures tail thickness and peak sharpness: leptokurtic distributions (>3 or excess kurtosis >0) have heavy tails and high outlier frequency compared to a normal distribution.
Statistical Reporting & Analysis Guidelines
- Sample (n-1) vs Population (N) Scope: Select 'Sample' if analyzing a subset of experimental subjects or polled data. Select 'Population' only if you possess the complete, exhaustive dataset of every existing member.
- Geometric Mean for Compounded Metrics: Use the Geometric Mean (nth root of product) rather than the Arithmetic Mean when averaging investment returns, ratios, or exponential bacterial growth rates.
- Standard Error of the Mean (SEM): Compute SEM = s / sqrt(n) to determine how precisely the sample mean estimates the true population mean for confidence interval modeling.
- Trimmed Mean for Heavy Tails: When datasets exhibit moderate outlier noise, a 5% or 10% Trimmed Mean (discarding top and bottom percentiles) provides balanced robustness.
- Coefficient of Variation (CV): Calculate CV = (s / Mean) * 100% to compare the relative dispersion of datasets with vastly different scales or measurement units.