Measures of Central Tendency

Arithmetic Mean - The Balancing Act

• Most common measure; the dataset's "center of gravity".
• Calculation: Sum of all values divided by the count of values.
  • Formula: $\bar{x} = \frac{\sum x_i}{n}$
  • Where $\sum x_i$ = sum of individual values, $n$ = number of values.
• Key Properties:
  • Simple to compute & interpret.
  • Includes all data points in calculation.
  • Highly affected by extreme values (outliers).
  • Sum of deviations of values from their mean is always 0: $\sum (x_i - \bar{x}) = 0$.
• Best for symmetrical data; less suitable for skewed distributions.

⭐ For a moderately skewed distribution, the empirical relationship is: Mean - Mode ≈ 3 (Mean - Median).

Median - The Resilient Middle

• Middlemost value in an ordered dataset (ascending/descending).
• Represents the 50th percentile; divides distribution into two equal halves.
• Calculation:
  • Odd (n) observations: The $(\frac{n+1}{2})^{th}$ value.
  • Even (n) observations: Average of the $(\frac{n}{2})^{th}$ and $(\frac{n}{2} + 1)^{th}$ values.
• Key advantage: Unaffected by extreme values (outliers), making it robust.
  • Preferred for skewed data (e.g., income, hospital stay duration, incubation period).
• 📌 "Median in the Middle" - stays central despite extreme pulls.

⭐ For skewed distributions (e.g., income data, incubation periods), Median is a more robust and often preferred measure of central tendency over Mean, as it is not influenced by outliers.

Mode - The Popular Peak

• Definition: The value that appears most frequently in a data set.
• Key Characteristics:
  • Represents the most common observation.
  • Can be:
    • Unimodal (one mode)
    • Bimodal (two modes)
    • Multimodal (>2 modes)
    • No mode (all values occur equally)
  • Unaffected by extreme values (outliers).
  • Only average for nominal (categorical) data.
• Empirical Relationship (moderately skewed distributions):
  • $Mode \approx 3 \cdot Median - 2 \cdot Mean$
• Advantages: Easy to understand; useful for qualitative data & identifying the most frequent size/category.
• Disadvantages: Not always unique/may not exist; not based on all values; poor for algebraic manipulation.

⭐ For a J-shaped distribution, the mode is at the highest point, which is at one end of the distribution.

Relationships & Selection - The Skewed Showdown

• Distribution & Central Tendency:
  • Symmetrical: Mean = Median = Mode.
  • Positively Skewed (Right): Mean > Median > Mode. 📌 Mean pulled by right tail.
  • Negatively Skewed (Left): Mean < Median < Mode. 📌 Mean pulled by left tail.
• Empirical Rule (Moderately Skewed, Unimodal):
  • $Mode \approx 3 \times Median - 2 \times Mean$.
• Selection Guide:
  • Mean: Best for normal quantitative data; sensitive to outliers.
  • Median: Best for skewed quantitative data, data with outliers, or ordinal data.
  • Mode: Best for nominal data; useful for bimodal/multimodal or qualitative data.

⭐ The Median is the most robust measure of central tendency for skewed distributions or data with extreme outliers as it is least affected by them.

High‑Yield Points - ⚡ Biggest Takeaways

• Mean (average) is most affected by outliers; use for normal distributions.
• Median (middle value, 50th percentile) is robust to outliers; best for skewed data.
• Mode is the most frequent value; useful for categorical/nominal data.
• Symmetrical distribution: Mean = Median = Mode.
• Positively skewed (right tail): Mean > Median > Mode.
• Negatively skewed (left tail): Mean < Median < Mode.
• Geometric Mean for growth rates/ratios; Harmonic Mean for average rates.