Quantitative Aptitude – Average
Level: MIT‑style advanced undergraduate (first principles to edge cases)
1. Introduction
The average (or mean) is a central notion in mathematics, statistics, and data science.
It is the simplest way to summarise a collection of numbers: one number that represents the “centre” of the collection.
In this note we:
- Derive the average from first principles.
- Explore algebraic properties and generalisations (weighted, continuous, vector‑valued).
- Connect the concept to probability theory (expectation).
- Discuss edge cases and pitfalls.
- Present a challenging problem that pushes the limits of your understanding.
Notation
- (x_1,x_2,\dots,x_n\in\mathbb{R}) – data points.
- ( \overline{x} ) – arithmetic mean.
- (w_i\ge 0) – weights.
- ( \mathbb{E}[X] ) – expectation of a random variable (X).
2. First Principles
2.1 Definition (Arithmetic Mean)
For a finite multiset (S={x_1,x_2,\dots,x_n}) with (n>0),
[ \boxed{\overline{x} ;=; \frac{1}{n}\sum_{i=1}^{n}x_i} ]
Why?
The mean is the unique point (m) that minimises the sum of squared deviations:
[ m=\arg\min_{y\in\mathbb{R}}\sum_{i=1}^{n}(x_i-y)^2 ]
Proof.
Expand the sum:
[ \sum_{i=1}^{n}(x_i-y)^2=\sum_{i=1}^{n}x_i^2-2y\sum_{i=1}^{n}x_i+ny^2 ]
Treat as a quadratic in (y); differentiate:
[ \frac{d}{dy}\left(\sum_{i=1}^{n}(x_i-y)^2\right)= -2\sum_{i=1}^{n}x_i+2ny ]
Set derivative to zero → (ny=\sum_{i=1}^{n}x_i) → (y=\frac{1}{n}\sum_{i=1}^{n}x_i).
Since the quadratic has a positive leading coefficient ((n>0)), this critical point is a global minimum.
2.2 Weighted Mean
Suppose each observation (x_i) is assigned a non‑negative weight (w_i) (with (\sum w_i>0)).
Define
[ \boxed{\overline{x}w ;=; \frac{\sum{i=1}^{n}w_ix_i}{\sum_{i=1}^{n}w_i}} ]
Derivation.
It is the minimiser of the weighted least‑squares problem:
[ \min_{y}\sum_{i=1}^{n}w_i(x_i-y)^2 ]
The same calculus gives (y=\overline{x}_w).
If all (w_i=1), we recover the ordinary mean.
3. Algebraic Properties
| Property | Statement | Proof Sketch |
|---|---|---|
| Linearity | (\overline{ax+by}=a\overline{x}+b\overline{y}) | Expand sums; constants factor out. |
| Monotonicity | If (x_i\le y_i) for all (i), then (\overline{x}\le\overline{y}) | Direct from sum inequality. |
| Translation | (\overline{x+c}=\overline{x}+c) | Add (c) inside sum; factor out. |
| Scale | (\overline{cx} = c\overline{x}) | Pull out (c). |
| Convexity | (\overline{x}\in [\min x_i,\max x_i]) | Weighted average of extremes. |
| Associativity | (\overline{\overline{x}y} = \overline{x}) when (x) and (y) have equal weights. | Two‑step averaging collapses. |
4. Continuous Analogues
4.1 Integral Mean (for a function)
Let (f:[a,b]\to\mathbb{R}) be integrable.
The average value of (f) over ([a,b]) is
[ \boxed{\langle f\rangle_{[a,b]}=\frac{1}{b-a}\int_{a}^{b}f(x),dx} ]
Connection to discrete mean.
If we partition ([a,b]) into (n) subintervals of equal length (\Delta x=(b-a)/n) and evaluate at midpoints (x_i), then
[ \frac{1}{n}\sum_{i=1}^{n}f(x_i)\approx\langle f\rangle_{[a,b]} ]
in the limit (n\to\infty).
4.2 Expectation of a Continuous Random Variable
If (X) has density (p(x)) on (\mathbb{R}), its expectation is precisely the weighted integral mean:
[ \mathbb{E}[X]=\int_{-\infty}^{\infty} x,p(x),dx ]
Thus average is the expected value of a random variable.
5. Statistical Interpretation
5.1 Sample Mean
Given a random sample (X_1,\dots,X_n) from a population with mean (\mu), the sample mean