Variance: Measures the spread of a distribution. Var(X) = E[(X - E[X])²] = E[X²] - (E[X])². Standard Deviation: σ = √Var(X).
- Calculate E[X] = Σx·P(X = x)
- Calculate E[X²] = Σx²·P(X = x)
- Apply the variance formula: Var(X) = E[X²] - (E[X])²
- Calculate SD(X) = √Var(X)
E[X] = (1)(0.2) + (2)(0.3) + (3)(0.4) + (4)(0.1) = 0.2 + 0.6 + 1.2 + 0.4 = 2.4
E[X²] = (1²)(0.2) + (2²)(0.3) + (3²)(0.4) + (4²)(0.1) = (1)(0.2) + (4)(0.3) + (9)(0.4) + (16)(0.1) = 0.2 + 1.2 + 3.6 + 1.6 = 6.6
Var(X) = E[X²] - (E[X])² = 6.6 - (2.4)² = 6.6 - 5.76 = 0.84
SD(X) = √Var(X) = √0.84 ≈ 0.9165
Var(X) = 0.84 and SD(X) ≈ 0.9165.
• Variance formula: Var(X) = E[X²] - (E[X])²
• Standard deviation: σ = √Var(X)
• Computational efficiency: Using the alternative variance formula
Continuous Variance: For a continuous random variable, Var(X) = ∫(x - E[X])²f(x)dx = E[X²] - (E[X])², where E[X] = ∫xf(x)dx and E[X²] = ∫x²f(x)dx.
E[X] = ∫₀¹ x·2x dx = ∫₀¹ 2x² dx = [2x³/3]₀¹ = (2/3)(1) - (2/3)(0) = 2/3
E[X²] = ∫₀¹ x²·2x dx = ∫₀¹ 2x³ dx = [2x⁴/4]₀¹ = (1/2)(1) - (1/2)(0) = 1/2
Var(X) = E[X²] - (E[X])² = 1/2 - (2/3)² = 1/2 - 4/9 = 9/18 - 8/18 = 1/18
Var(X) = 1/18 and SD(X) = √(1/18) ≈ 0.2357.
• Continuous variance: Var(X) = E[X²] - (E[X])²
• Integration: Apply power rule for polynomials
• Domain: Integrate only over where f(x) > 0
Variance Properties: For constants a and b: Var(aX + b) = a²Var(X). The constant b does not affect variance. SD(aX + b) = |a|SD(X).
For any constants a and b: Var(aX + b) = a²Var(X). The constant b shifts the distribution but doesn't affect spread.
For 3X + 2: a = 3 and b = 2.
Var(3X + 2) = 3²Var(X) = 9 × 5 = 45
SD(3X + 2) = √Var(3X + 2) = √45 = 3√5 ≈ 6.708
Var(3X + 2) = 45 and SD(3X + 2) = 3√5 ≈ 6.708.
• Variance scaling: Var(aX + b) = a²Var(X)
• Constant shift: Adding a constant doesn't change variance
• Standard deviation scaling: SD(aX + b) = |a|SD(X)
Variance: A measure of how spread out the values of a random variable are from the mean. Var(X) = E[(X - E[X])²].
Standard Deviation: The square root of variance, giving the average deviation from the mean in the same units as the random variable.
Discrete Variance: Var(X) = Σ(x - μ)²·P(X = x) where μ = E[X].
Continuous Variance: Var(X) = ∫(x - μ)²f(x)dx where μ = E[X] and f(x) is the PDF.
Alternative Formula: Var(X) = E[X²] - (E[X])², often computationally more efficient.
- Calculate E[X]: Find the expected value
- Calculate E[X²]: Find the expected value of X²
- Apply variance formula: Var(X) = E[X²] - (E[X])²
- Find standard deviation: SD(X) = √Var(X)
- Apply properties: Use scaling properties when applicable
- Interpret results: Understand practical significance
• Variance: Var(X) = E[(X - E[X])²] = E[X²] - (E[X])²
• Standard deviation: SD(X) = √Var(X)
• Scaling: Var(aX + b) = a²Var(X)
• Sum (independent): Var(X + Y) = Var(X) + Var(Y)
• Discrete: Var(X) = Σ(x - μ)²·P(X = x)
• Continuous: Var(X) = ∫(x - μ)²f(x)dx
Binomial Distribution: X ~ B(n,p) where n is the number of trials and p is the probability of success. For binomial: E[X] = np and Var(X) = np(1-p).
Flipping a fair coin 10 times follows a binomial distribution: X ~ B(10, 0.5).
For X ~ B(n,p): Var(X) = np(1-p) = 10 × (0.5) × (1-0.5) = 10 × 0.5 × 0.5 = 2.5
SD(X) = √Var(X) = √2.5 ≈ 1.581
Var(X) = 2.5 and SD(X) ≈ 1.581.
• Binomial variance: Var(X) = np(1-p)
• Standard deviation: SD(X) = √Var(X)
• Known distribution: Use specialized formulas when available
Independent Random Variables: If X and Y are independent, then Var(X + Y) = Var(X) + Var(Y). For linear combinations: Var(aX + bY + c) = a²Var(X) + b²Var(Y).
For independent X and Y: Var(aX + bY + c) = a²Var(X) + b²Var(Y). The constant c doesn't affect variance.
For 2X - 3Y + 5: a = 2, b = -3, and c = 5.
Var(2X - 3Y + 5) = (2)²Var(X) + (-3)²Var(Y) = 4×4 + 9×9 = 16 + 81 = 97
Var(2X - 3Y + 5) = 97.
• Independence property: Var(X + Y) = Var(X) + Var(Y) for independent variables
• Linear combination: Var(aX + bY + c) = a²Var(X) + b²Var(Y)
• Constant addition: Adding a constant doesn't change variance
Variance: A measure of how spread out the values of a random variable are from the mean. It quantifies the dispersion of the distribution.
Standard Deviation: The square root of variance, providing the average deviation from the mean in the same units as the random variable.
Discrete Variance: Var(X) = Σ(x - μ)²·P(X = x) where μ = E[X].
Continuous Variance: Var(X) = ∫(x - μ)²f(x)dx where μ = E[X] and f(x) is the PDF.
Alternative Formula: Var(X) = E[X²] - (E[X])², often computationally more efficient.
Properties: Var(aX + b) = a²Var(X) for constants a and b.
- Calculate E[X]: Find the expected value of the random variable
- Calculate E[X²]: Find the expected value of X²
- Apply variance formula: Use Var(X) = E[X²] - (E[X])²
- Find standard deviation: Calculate SD(X) = √Var(X)
- Apply properties: Use scaling and transformation properties when applicable
- Interpret results: Understand the practical significance of the measures
• Variance: Var(X) = E[(X - E[X])²] = E[X²] - (E[X])²
• Standard deviation: SD(X) = √Var(X)
• Scaling: Var(aX + b) = a²Var(X)
• Sum (independent): Var(X + Y) = Var(X) + Var(Y)
• Linear combination (independent): Var(aX + bY + c) = a²Var(X) + b²Var(Y)
• Discrete: Var(X) = Σ(x - μ)²·P(X = x)
• Continuous: Var(X) = ∫(x - μ)²f(x)dx
• Binomial: Var(X) = np(1-p) for X ~ B(n,p)
• Geometric: Var(X) = (1-p)/p² for X ~ Geo(p)
• Simple: Fair die roll (variance calculation)
• Intermediate: Given probability distribution
• Advanced: Continuous probability density function
• Complex: Linear combinations of random variables
• Variance measures how spread out values are from the mean
• Standard deviation is in the same units as the original variable
• Var(aX + b) = a²Var(X) - constant doesn't affect variance
• Use E[X²] - (E[X])² for computational efficiency
• For independent variables, variances add: Var(X + Y) = Var(X) + Var(Y)
Questions & Answers
Question: Why do we square the deviations in variance calculation instead of using absolute values?
Answer: Squaring the deviations provides several advantages:
- Mathematical convenience: Squares are differentiable, making optimization easier
- Penalizing outliers: Larger deviations are penalized more heavily
- Algebraic properties: Leads to nice properties like Var(X + Y) = Var(X) + Var(Y) for independent variables
- Connection to normal distribution: Many statistical methods assume normality, which is naturally expressed in terms of squared deviations
While absolute deviations would work, squaring leads to more elegant mathematical properties.
Question: Why does adding a constant not change the variance?
Answer: Adding a constant shifts the entire distribution by the same amount, but doesn't change the spread:
- Visual analogy: Imagine moving a seesaw - the balance point changes, but the distances from the center stay the same
- Mathematical reason: Var(X + c) = E[(X + c - E[X + c])²] = E[(X + c - (E[X] + c))²] = E[(X - E[X])²] = Var(X)
- Practical example: If everyone in a class scores 5 points higher, the average changes but the spread of scores stays the same
Variance measures spread, not location.
Question: When do we add variances of independent random variables versus when we don't?
Answer: Var(X + Y) = Var(X) + Var(Y) holds only when X and Y are independent (or uncorrelated):
- Independent variables: Var(X + Y) = Var(X) + Var(Y)
- Dependent variables: Var(X + Y) = Var(X) + Var(Y) + 2Cov(X,Y)
- General rule: Var(aX + bY) = a²Var(X) + b²Var(Y) + 2abCov(X,Y)
The independence assumption is crucial for the simple addition formula. If variables are correlated, you must account for their covariance.