Back Matter — Overview

Appendices at a Glance

The appendices that follow support every chapter of this textbook. They provide mathematical prerequisites, formula sheets, distribution references, computing guides, a capstone rubric, and a readiness check for advanced study. Use them for review, reference, and transition into mathematical statistics, machine learning, econometrics, and quantitative research.

20Appendices

80+Key Terms

17Distributions

6Readiness Sections

App ACalculus Readiness ReviewLimits, differentiation, integration, double integrals, series
App BIntegration Techniques for ProbabilityImproper integrals, integration by parts, gamma function, beta function, Gaussian integral
App CLinear Algebra Readiness ReviewMatrices, transpose, inverse, determinants, eigenvalues — supporting regression and multivariate theory
App DSet and Event Notation ReferenceUnion, intersection, complement, subset, partition, disjoint events, set-builder notation
App EProbability Rules ReferenceKolmogorov axioms, complement, addition, multiplication, conditional, total probability, Bayes' theorem, independence
App FCommon Discrete Distributions ReferenceBernoulli, Binomial, Geometric, Negative Binomial, Hypergeometric, Poisson — PMF, mean, variance, MGF
App GCommon Continuous Distributions ReferenceUniform, Exponential, Normal, Gamma, Beta, Chi-Squared, Student's t, F — PDF, CDF, mean, variance, MGF
App HExpected Value and Variance Formula SheetDefinitions, linearity, properties, moments, skewness, kurtosis
App IJoint Distribution ReferenceJoint PMF and PDF, marginals, conditional distributions, independence, covariance, correlation
App JMoment Generating Function ReferenceDefinition, moments, uniqueness, independence, linear transformation, MGF table
App KLimit Theorems ReferenceLaw of large numbers (weak and strong), central limit theorem, normal approximation to binomial, standard error
App LSampling Distribution ReferenceSample mean, sample proportion, sample variance, key sampling distributions under normality
App MEstimation and Confidence Interval ReferencePoint estimation, bias, MSE, confidence interval formulas, critical values, interpretation
App NHypothesis Testing ReferenceTest structure, error types, power, common test statistics and null distributions, p-value interpretation
App OLikelihood and Bayesian ReferenceLikelihood function, log-likelihood, MLE table, Bayesian framework, conjugate priors, point estimates
App PRegression and ANOVA ReferenceSimple and multiple regression formulas, one-way ANOVA table, assumptions
App QStochastic Process Preview ReferenceBernoulli process, random walk, Markov chains, transition matrix, stationary distribution, Poisson process
App RStatistical Computing and Technology GuideR, Python, graphing calculators — code for all major distributions, tests, and regression
App SCapstone Project RubricResearch question, sampling, EDA, model selection, inference, interpretation, communication, ethics
App TAdvanced Statistics and Data Science Readiness CheckSix sections covering probability, joint distributions, estimation, regression, Bayesian reasoning, stochastic processes

Appendix A

Calculus Readiness Review

A.1 Limits and Continuity

A limit describes the value a function approaches as the input approaches a given point. For probability, limits arise in defining density functions, verifying that total probability equals one, and understanding convergence in the law of large numbers and central limit theorem.

Key Results

If f is continuous at a, then lim(x → a) f(x) = f(a).

Squeeze theorem: if g(x) ≤ f(x) ≤ h(x) near a and lim g(x) = lim h(x) = L, then lim f(x) = L.

A.2 Differentiation Review

The derivative f′(x) measures the instantaneous rate of change of f at x. In probability, derivatives appear when converting between cumulative distribution functions and probability density functions: if F(x) is the CDF of a continuous random variable, then f(x) = F′(x) is the PDF.

Rule	Formula
Power rule	`d/dx [xⁿ] = nxⁿ⁻¹`
Product rule	`d/dx [uv] = u′v + uv′`
Chain rule	`d/dx [f(g(x))] = f′(g(x)) · g′(x)`
Exponential	`d/dx [eˣ] = eˣ; d/dx [eᵃˣ] = aeᵃˣ`
Natural logarithm	`d/dx [ln x] = 1/x`

A.3 Integration Review

The definite integral ∫ₐᵇ f(x) dx measures the signed area under f between a and b. In probability, probability is area under a density curve, expected value is a weighted integral, and variance is computed by integrating a squared deviation.

Technique	Formula
Power rule	`∫ xⁿ dx = xⁿ⁺¹/(n+1) + C, n ≠ −1`
Exponential	`∫ eˣ dx = eˣ + C; ∫ eᵃˣ dx = eᵃˣ/a + C`
Logarithm	`∫ (1/x) dx = ln\|x\| + C`
Substitution	Let u = g(x), du = g′(x) dx
Integration by parts	`∫ u dv = uv − ∫ v du`
Improper integrals	`∫ₐ^∞ f(x) dx = lim(b → ∞) ∫ₐᵇ f(x) dx`

A.4 Double Integrals (Preview)

Joint distributions of two continuous random variables require integration over a two-dimensional region. The double integral ∬_R f(x, y) dA computes the total probability over a region R. Fubini's theorem permits evaluation as iterated integrals when f is continuous:

∬_R f(x, y) dA = ∫ₐᵇ [∫_c^d f(x, y) dy] dx

A.5 Series and Summation Review

Discrete probability uses infinite sums. A geometric series with ratio r, |r| < 1:

Σₙ₌₀^∞ rⁿ = 1/(1 − r)

The Taylor series for eˣ, which underlies the Poisson distribution and moment generating functions:

eˣ = Σₙ₌₀^∞ xⁿ/n!

A.6Calculus Readiness Problems

Evaluate ∫₀^∞ e⁻²ˣ dx.
Differentiate F(x) = 1 − e⁻³ˣ for x > 0. Interpret the result in terms of a PDF.
Verify that ∫₋∞^∞ (1/√(2π)) e^(−x²/2) dx = 1 requires only knowing the Gaussian integral result.
Evaluate ∫₀¹ x(1 − x) dx.
Use integration by parts to evaluate ∫₀^∞ x e⁻ˣ dx.
Evaluate the double integral ∫₀¹ ∫₀¹ (x + y) dy dx.

Appendix B

Integration Techniques for Probability

B.1 Improper Integrals Over Infinite Domains

Many probability density functions are defined on [0, ∞) or (−∞, ∞). If the limit is finite, the integral converges. A density function must satisfy ∫ f(x) dx = 1 over its entire domain.

Example B.1

Verifying the exponential density integrates to 1

Givenf(x) = λe^(−λx) for x ≥ 0, λ > 0
Solve∫₀^∞ λe^(−λx) dx = λ · [−e^(−λx)/λ]₀^∞ = λ · (1/λ) = 1 ✓

B.2 Integration by Parts in Expected Value

Example B.2

E[X] for the exponential distribution

SetupE[X] = ∫₀^∞ x λe^(−λx) dx. Let u = x, dv = λe^(−λx) dx.
Thendu = dx, v = −e^(−λx)
SolveE[X] = [−xe^(−λx)]₀^∞ + ∫₀^∞ e^(−λx) dx = 0 + 1/λ = 1/λ

B.3 Gamma Function

The gamma function is defined for α > 0 as Γ(α) = ∫₀^∞ xᵅ⁻¹ e⁻ˣ dx. It appears in the gamma distribution, beta distribution, normal distribution normalization, and chi-squared distribution.

Property	Statement
Base case	`Γ(1) = 1`
Recursion	`Γ(α) = (α − 1)Γ(α − 1) for α > 1`
Integer values	`Γ(n) = (n − 1)! for positive integers n`
Half-integer	`Γ(1/2) = √π`

B.4 Beta Function

B(α, β) = ∫₀¹ xᵅ⁻¹ (1 − x)^(β−1) dx = Γ(α)Γ(β)/Γ(α + β)

Connection

The beta function normalizes the beta distribution PDF and appears in Bayesian inference when computing posteriors with Beta conjugate priors.

B.5 Gaussian Integral

The key result underlying the normal distribution normalization:

∫₋∞^∞ e^(−x²) dx = √π Therefore: ∫₋∞^∞ e^(−x²/2) dx = √(2π)

B.6Multivariate Integration Practice

Verify that f(x, y) = 6xy for 0 < x < 1, 0 < y < 1, 0 < x + y < 1 is a valid joint density, or state why it is not.
Evaluate E[X] when f(x, y) = 2 for 0 < x < y < 1.
Use the gamma function to show that the gamma distribution with parameters α and β has mean αβ.
Evaluate ∫₀^∞ x³ e^(−x) dx using the gamma function.

Appendix C

Linear Algebra Readiness Review

C.1 Vectors and Matrices

A matrix is a rectangular array of numbers. An m × n matrix has m rows and n columns. In regression and multivariate analysis, data and parameters are organized in matrices. Matrix multiplication: the (i, j) entry of AB is the dot product of row i of A and column j of B.

C.2 Transpose and Symmetric Matrices

The transpose Aᵀ swaps rows and columns. A matrix is symmetric if A = Aᵀ. Covariance matrices in multivariate probability are always symmetric and positive semidefinite.

C.3 Matrix Inverse

A is invertible if AA⁻¹ = A⁻¹A = I. In multiple regression, the ordinary least squares formula is:

β̂ = (XᵀX)⁻¹ Xᵀy, provided XᵀX is invertible

C.4 Determinants

det(A) ≠ 0 if and only if A is invertible. Determinants appear in multivariate change-of-variable formulas through the Jacobian.

C.5 Eigenvalues and Eigenvectors (Preview)

A scalar λ and nonzero vector v satisfy Av = λv if λ is an eigenvalue and v is the corresponding eigenvector. Eigenvalues of covariance matrices describe the spread in principal directions and are central to principal component analysis.

C.6Linear Algebra Readiness Problems

Multiply A = [[2, 1], [0, 3]] and B = [[1, −1], [4, 2]].
Find the inverse of A = [[3, 1], [5, 2]].
Compute det([[4, 7], [2, 6]]).
Write the normal equations XᵀXβ = Xᵀy for a simple linear regression with n = 3 observations.
If Σ is a 2×2 symmetric matrix with eigenvalues 4 and 1, describe what this means geometrically for a bivariate distribution.

Appendices D – E

Notation and Probability Rules Reference

Appendix D: Set and Event Notation

Symbol	Meaning	Example
`Ω`	Sample space; set of all outcomes	Ω = {H, T} for a coin flip
`∅`	Empty set; impossible event	P(∅) = 0
`A ∪ B`	Union; A or B or both occur	{x : x ∈ A or x ∈ B}
`A ∩ B`	Intersection; both A and B occur	{x : x ∈ A and x ∈ B}
`Aᶜ`	Complement of A; A does not occur	P(Aᶜ) = 1 − P(A)
`A ⊆ B`	A is a subset of B	If A occurs, B must occur
`A \ B`	Set difference; in A but not B	A \ B = A ∩ Bᶜ
`\|A\|`	Cardinality; number of elements in A	\|{1,2,3}\| = 3
`x ∈ A`	x is an element of A	2 ∈ {1, 2, 3}
`{x : condition}`	Set-builder notation	{x : x > 0}

Three Key Concepts

Disjoint (mutually exclusive): A ∩ B = ∅. They cannot both occur.

Exhaustive: A₁ ∪ A₂ ∪ ··· ∪ Aₙ = Ω. Together they cover all outcomes.

Partition: A collection {A₁, …, Aₙ} that is both pairwise disjoint and exhaustive.

Appendix E: Probability Rules Reference

Kolmogorov Axioms: (1) P(A) ≥ 0 for every event A. (2) P(Ω) = 1. (3) If A₁, A₂, … are pairwise disjoint, then P(A₁ ∪ A₂ ∪ ···) = P(A₁) + P(A₂) + ···

Rule	Formula
Complement	`P(Aᶜ) = 1 − P(A)`
Addition (General)	`P(A ∪ B) = P(A) + P(B) − P(A ∩ B)`
Addition (Mutually Exclusive)	`P(A ∪ B) = P(A) + P(B)`
Multiplication (General)	`P(A ∩ B) = P(A) · P(B \| A)`
Multiplication (Independent)	`P(A ∩ B) = P(A) · P(B)`
Conditional Probability	`P(B \| A) = P(A ∩ B) / P(A), P(A) > 0`
Law of Total Probability	`P(B) = Σᵢ P(B \| Aᵢ)P(Aᵢ) for a partition {Aᵢ}`
Bayes' Theorem	`P(Aᵢ \| B) = P(B \| Aᵢ)P(Aᵢ) / Σⱼ P(B \| Aⱼ)P(Aⱼ)`

Independence

Events A and B are independent if and only if P(A ∩ B) = P(A) · P(B). A collection of events A₁, …, Aₙ is mutually independent if the multiplication rule holds for every subcollection.

Appendices F – G

Distribution Reference Tables

Distribution	PMF P(X = x)	Mean	Variance	MGF Mₓ(t)
Bernoulli(p) `x ∈ {0, 1}`	`P(1) = p, P(0) = 1−p`	`p`	`p(1−p)`	`1−p+peᵗ`
Binomial(n, p) `x ∈ {0,…,n}`	`C(n,x) pˣ (1−p)ⁿ⁻ˣ`	`np`	`np(1−p)`	`(1−p+peᵗ)ⁿ`
Geometric(p) `x ∈ {1,2,3,…}`	`(1−p)ˣ⁻¹ p`	`1/p`	`(1−p)/p²`	`peᵗ/(1−(1−p)eᵗ)`
Neg. Binomial(r, p) `x ∈ {r, r+1, …}`	`C(x−1,r−1) pʳ (1−p)ˣ⁻ʳ`	`r/p`	`r(1−p)/p²`	—
Hypergeometric(N,K,n)	`C(K,x)C(N−K,n−x)/C(N,n)`	`nK/N`	`n(K/N)(1−K/N)(N−n)/(N−1)`	—
Poisson(λ) `x ∈ {0,1,2,…}`	`e^(−λ) λˣ / x!`	`λ`	`λ`	`exp(λ(eᵗ−1))`

Distribution	PDF f(x)	Mean	Variance	Key Property
Uniform(a, b) `x ∈ [a, b]`	`1/(b−a)`	`(a+b)/2`	`(b−a)²/12`	Equal density everywhere on [a, b]
Exponential(λ) `x ∈ [0, ∞)`	`λe^(−λx)`	`1/λ`	`1/λ²`	Memoryless: P(X>s+t\|X>s) = P(X>t)
Normal(μ, σ²) `x ∈ (−∞, ∞)`	`(1/σ√(2π)) exp(−(x−μ)²/2σ²)`	`μ`	`σ²`	Standard form: Z = (X−μ)/σ ~ N(0,1)
Gamma(α, β) `x ∈ [0, ∞)`	`xᵅ⁻¹ e^(−x/β) / (βᵅ Γ(α))`	`αβ`	`αβ²`	Exp(λ) = Gamma(1, 1/λ); χ²(k) = Gamma(k/2, 2)
Beta(α, β) `x ∈ [0, 1]`	`xᵅ⁻¹(1−x)^(β−1) / B(α,β)`	`α/(α+β)`	`αβ/((α+β)²(α+β+1))`	Models proportions; conjugate prior for Binomial
Chi-Squared(k) `x ∈ [0, ∞)`	—	`k`	`2k`	X = Z₁²+···+Zₖ², Zᵢ ~ N(0,1) indep.
Student's t(k) `x ∈ (−∞, ∞)`	—	0 (k>1)	`k/(k−2) (k>2)`	T = Z/√(V/k); approaches N(0,1) as k→∞
F(d₁, d₂) `x ∈ (0, ∞)`	—	`d₂/(d₂−2)`	—	Ratio of two χ² variables divided by df; used in ANOVA and regression

Appendices H – L

Expected Value, Moments, Joints, MGFs, and Limit Theorems

Appendix H: Expected Value and Variance Formula Sheet

Quantity	Discrete	Continuous
E[X]	`Σₓ x · p(x)`	`∫₋∞^∞ x f(x) dx`
E[g(X)]	`Σₓ g(x) · p(x)`	`∫₋∞^∞ g(x) f(x) dx`
Var(X)	`E[(X−μ)²] = E[X²] − (E[X])²`

Property	Formula
Linearity	`E[aX + b] = aE[X] + b`
Variance of linear transformation	`Var(aX + b) = a² Var(X)`
Sum of independent variables	`E[X + Y] = E[X] + E[Y]`
Variance of sum (independent)	`Var(X + Y) = Var(X) + Var(Y)`
Variance of sum (dependent)	`Var(X + Y) = Var(X) + Var(Y) + 2 Cov(X, Y)`
Skewness	`γ₁ = μ₃/σ³`
Kurtosis (excess)	`γ₂ − 3, where γ₂ = μ₄/σ⁴`

Appendix I: Joint Distribution Reference

Concept	Discrete Formula	Continuous Formula
Marginal of X	`pₓ(x) = Σᵧ p(x,y)`	`fₓ(x) = ∫₋∞^∞ f(x,y) dy`
Conditional Y\|X	`p(y\|x) = p(x,y)/pₓ(x)`	`f(y\|x) = f(x,y)/fₓ(x)`
Independence	`f(x,y) = fₓ(x) · fᵧ(y) for all (x,y)`
Covariance	`Cov(X,Y) = E[XY] − E[X]E[Y]`
Correlation	`ρ(X,Y) = Cov(X,Y) / (SD(X) · SD(Y)), ρ ∈ [−1,1]`

Critical Warning

ρ = 0 does not imply independence in general. Independence implies ρ = 0, but the converse is false. Always verify independence through the joint factorization condition, not correlation alone.

Appendix J: Moment Generating Function Reference

Definition: Mₓ(t) = E[eᵗˣ], defined for all t in an open interval containing 0.

Property	Statement
Moments	`E[Xᵏ] = Mₓ⁽ᵏ⁾(0) (kth derivative at t = 0)`
Uniqueness	Same MGF in interval around 0 → identical distributions
Independence	`M_{X+Y}(t) = Mₓ(t) · Mᵧ(t) if X, Y independent`
Linear transformation	`M_{aX+b}(t) = eᵇᵗ Mₓ(at)`

Distribution	MGF Mₓ(t)
Bernoulli(p)	`1 − p + peᵗ`
Binomial(n, p)	`(1 − p + peᵗ)ⁿ`
Poisson(λ)	`exp(λ(eᵗ − 1))`
Exponential(λ)	`λ/(λ − t), t < λ`
Normal(μ, σ²)	`exp(μt + σ²t²/2)`
Gamma(α, β)	`(1 − βt)^(−α), t < 1/β`

Appendix K: Limit Theorems Reference

Law of Large Numbers

Let X₁, X₂, … be i.i.d. with mean μ and finite variance. Then for every ε > 0:

lim(n → ∞) P(|X̄ₙ − μ| > ε) = 0

The sample mean converges in probability to the population mean.

Central Limit Theorem

Let X₁, X₂, … be i.i.d. with mean μ and finite variance σ². Then as n → ∞:

(X̄ₙ − μ) / (σ/√n) →^d N(0, 1)

X̄ₙ is approximately N(μ, σ²/n) for large n. Practical guidance: n ≥ 30 is often sufficient for approximate normality when the population is not severely skewed or heavy-tailed.

Normal Approximation to the Binomial

If X ~ Binomial(n, p) and np ≥ 10 and n(1−p) ≥ 10, then X ≈ N(np, np(1−p)).

Continuity correction: P(X ≤ k) ≈ Φ((k + 0.5 − np)/√(np(1−p)))

Appendix L: Sampling Distribution Reference

Statistic	Mean	Variance / SE	Exact Distribution (Normal Pop.)
X̄ (σ known)	μ	`σ²/n; SE = σ/√n`	`(X̄−μ)/(σ/√n) ~ N(0,1)`
X̄ (σ unknown)	μ	`σ²/n`	`(X̄−μ)/(S/√n) ~ t(n−1)`
p̂ = X/n	p	`p(1−p)/n; SE = √(p(1−p)/n)`	CLT: approx. N(p, p(1−p)/n)
(n−1)S²/σ²	n−1	—	`χ²(n−1)`
S₁²/S₂² · (σ₂²/σ₁²)	—	—	`F(n₁−1, n₂−1)`

Denominator of Sample Variance

The sample variance S² = (1/(n−1)) Σ(Xᵢ − X̄)² uses n − 1, not n, to correct for using X̄ instead of the unknown μ. This makes S² an unbiased estimator: E[S²] = σ².

Appendices M – O

Estimation, Hypothesis Testing, and Likelihood

Appendix M: Estimation and Confidence Interval Reference

Property	Definition
Bias	`Bias(T) = E[T] − θ`
Unbiased	`E[T] = θ`
MSE	`MSE(T) = Var(T) + [Bias(T)]²`
Consistent	`T →^p θ as n → ∞`

Parameter	Conditions	Confidence Interval
μ (σ known)	CLT or normal pop.	`X̄ ± z*α/2 · σ/√n`
μ (σ unknown)	Normal pop. or large n	`X̄ ± t*α/2(n−1) · S/√n`
p	np̂ ≥ 10, n(1−p̂) ≥ 10	`p̂ ± z*α/2 · √(p̂(1−p̂)/n)`
μ₁ − μ₂ (indep.)	σ₁, σ₂ unknown; large n	`(X̄₁ − X̄₂) ± t* · √(S₁²/n₁ + S₂²/n₂)`

Confidence Level	Critical Value z*
90%	1.645
95%	1.960
99%	2.576

Interpreting Confidence Intervals

A 95% CI does not mean the true parameter is in this specific interval with probability 95%. It means the method produces intervals containing the true parameter in 95% of repeated applications under the stated assumptions.

Appendix N: Hypothesis Testing Reference

Structure of a Hypothesis Test

1. State hypotheses: H₀ (null) versus H₁ (alternative).

2. Choose significance level α (common values: 0.01, 0.05, 0.10).

3. Identify test statistic and its distribution under H₀.

4. Compute the observed test statistic from the data.

5. Compute the p-value: P(result at least as extreme as observed | H₀ true).

6. Decision: reject H₀ if p-value < α.

7. State conclusion in context.

Decision	H₀ True	H₀ False
Reject H₀	Type I Error (prob. = α)	Correct (power = 1 − β)
Fail to Reject H₀	Correct (prob. = 1 − α)	Type II Error (prob. = β)

Test	Statistic	Null Distribution
One-sample mean (σ known)	`z = (X̄ − μ₀)/(σ/√n)`	N(0,1)
One-sample mean (σ unknown)	`t = (X̄ − μ₀)/(S/√n)`	t(n−1)
One-sample proportion	`z = (p̂ − p₀)/√(p₀(1−p₀)/n)`	N(0,1)
Two-sample means	`t = (X̄₁ − X̄₂)/√(S₁²/n₁ + S₂²/n₂)`	t (Welch df)
Variance	`χ² = (n−1)S²/σ₀²`	χ²(n−1)
ANOVA	`F = MSB/MSW`	F(k−1, N−k)

p-Value Interpretation

The p-value is not the probability that H₀ is true. It is the probability, under H₀, of a test statistic at least as extreme as observed. A small p-value indicates that the data would be surprising if H₀ were true. Statistical significance does not imply practical importance.

Appendix O: Likelihood and Bayesian Reference

For observed data x = (x₁, …, xₙ) and parameter θ, the likelihood function is:

L(θ; x) = f(x₁; θ) · f(x₂; θ) · ··· · f(xₙ; θ)

The log-likelihood is ℓ(θ; x) = ln L(θ; x). The MLE is θ̂ = argmax L(θ; x).

Distribution	MLE
Bernoulli(p)	`p̂ = x̄`
Poisson(λ)	`λ̂ = x̄`
Normal(μ, σ²)	`μ̂ = x̄, σ̂² = (1/n)Σ(xᵢ − x̄)²`
Exponential(λ)	`λ̂ = 1/x̄`

Bayesian Framework

Prior: π(θ) — beliefs about θ before seeing data.

Posterior: π(θ | x) ∝ L(θ; x) · π(θ)

In words: posterior ∝ likelihood × prior

Likelihood	Conjugate Prior	Posterior
Binomial(n, p)	Beta(α, β)	`Beta(α + x, β + n − x)`
Poisson(λ)	Gamma(α, β)	`Gamma(α + Σxᵢ, β + n)`
Normal(μ, σ² known)	Normal(μ₀, τ₀²)	Normal (updated mean and variance)

Appendices P – Q

Regression, ANOVA, and Stochastic Processes

Appendix P: Regression and ANOVA Reference

Simple Linear Regression Model: Yᵢ = β₀ + β₁xᵢ + εᵢ, where εᵢ ~ i.i.d. N(0, σ²)

Quantity	Formula
Slope estimate	`β̂₁ = Σ(xᵢ − x̄)(yᵢ − ȳ) / Σ(xᵢ − x̄)²`
Intercept estimate	`β̂₀ = ȳ − β̂₁x̄`
Fitted values	`ŷᵢ = β̂₀ + β̂₁xᵢ`
Residuals	`eᵢ = yᵢ − ŷᵢ`
R²	`1 − SSE/SST, SST = Σ(yᵢ−ȳ)², SSE = Σeᵢ²`
SE of slope	`SE(β̂₁) = s / √Σ(xᵢ−x̄)², s² = SSE/(n−2)`
t-test for slope	`t = β̂₁/SE(β̂₁) ~ t(n−2) under H₀: β₁ = 0`

Multiple Regression in Matrix Form: Y = Xβ + ε. Least squares solution: β̂ = (XᵀX)⁻¹ Xᵀy.

One-Way ANOVA Table

Source	SS	df	MS	F
Between groups	`SSB = Σnᵢ(ȳᵢ−ȳ)²`	k−1	MSB = SSB/(k−1)	MSB/MSW
Within groups	`SSW = ΣΣ(yᵢⱼ−ȳᵢ)²`	N−k	MSW = SSW/(N−k)	—
Total	SST	N−1	—	—

ANOVA Assumptions

1. Independence of observations. 2. Normality within each group (or large n for CLT). 3. Equal variances across groups (homoscedasticity). Always verify assumptions before interpreting F-test results.

Appendix Q: Stochastic Process Preview Reference

Process	Key Properties	Distribution Result
Bernoulli Process	i.i.d. Bernoulli(p) trials	Successes in n trials ~ Binomial(n, p); waiting time ~ Geometric(p)
Random Walk Sₙ = X₁+···+Xₙ	Xᵢ = +1 (prob. p), −1 (prob. 1−p); i.i.d.	E[Sₙ] = n(2p−1); Var(Sₙ) = 4p(1−p)n
Markov Chain	Markov property: future\|present ⊥ past	n-step: P^n; stationary π: πP = π, Σπⱼ = 1
Poisson Process N(t), rate λ	N(0)=0; indep. increments; N(t+s)−N(t)~Poisson(λs)	Interarrivals ~ Exp(λ); kth arrival ~ Gamma(k, 1/λ)

Appendix R

Statistical Computing and Technology Guide

R — Essential Commands

# Probability distributions: d (density/PMF), p (CDF), q (quantile), r (random sample)
dbinom(x, size, prob)        # Binomial PMF
pbinom(q, size, prob)        # Binomial CDF
qnorm(p, mean, sd)           # Normal quantile (critical values)
pnorm(q, mean, sd)           # Normal CDF (area to the left)
rnorm(n, mean, sd)           # Normal random sample
# Also: dpois, ppois; dexp, pexp; dt, pt, qt; dchisq, pchisq, qchisq; df, pf, qf

# Descriptive statistics
mean(x); median(x); var(x); sd(x); quantile(x, probs)

# Hypothesis tests
t.test(x, mu = 0)                   # One-sample t-test
t.test(x, y)                         # Two-sample t-test
prop.test(x, n, p = 0.5)            # Proportion test
chisq.test(x)                        # Chi-squared test
aov(response ~ factor, data = df)    # One-way ANOVA

# Linear regression
model <- lm(y ~ x, data = df)
summary(model)                        # Coefficients, R², F-test, p-values
model2 <- lm(y ~ x1 + x2, data = df) # Multiple regression

Python — Essential Commands

from scipy import stats
import numpy as np
import statsmodels.api as sm

# Normal distribution
stats.norm.pdf(x, loc=0, scale=1)     # PDF
stats.norm.cdf(x, loc=0, scale=1)     # CDF
stats.norm.ppf(p, loc=0, scale=1)     # Quantile (inverse CDF)

# Hypothesis tests
t_stat, p_val = stats.ttest_1samp(data, popmean=0)
t_stat, p_val = stats.ttest_ind(group1, group2)
chi2, p_val   = stats.chisquare(observed, expected)

# Regression with statsmodels
X = sm.add_constant(X)
model = sm.OLS(y, X).fit()
print(model.summary())

# Simulation
np.random.seed(42)
samples = np.random.binomial(n=10, p=0.3, size=1000)
means   = np.random.normal(loc=5, scale=2, size=500)

Graphing Calculators (TI-84 and similar)

Function	Location	Purpose
`normalcdf(L, U, μ, σ)`	2nd DISTR	Normal probability P(L < X < U)
`invNorm(area, μ, σ)`	2nd DISTR	Normal critical value
`binomcdf, binompdf`	2nd DISTR	Binomial CDF and PMF
`1-Var Stats`	STAT → CALC	Descriptive statistics
`T-Test, Z-Test`	STAT → TESTS	Hypothesis tests for means
`1-PropZTest`	STAT → TESTS	Proportion test

Using Technology Responsibly

State the model and assumptions before computing
Verify that assumptions are reasonable for the data and context
Report results with units and in context
Assess practical importance, not only statistical significance
Recognize when output is unreliable (small samples, violated assumptions, extreme outliers)

Appendix S

Capstone Project Rubric

The capstone project in Chapter 14 asks you to carry out a complete statistical analysis from data collection through written communication. The rubric below describes expectations at each level. Total: 100 points.

1. Research Question and Context — 10 points

Score	Description
9–10	Question is specific, testable, and clearly motivated. Context is explained precisely. Population of interest is defined.
7–8	Question is clear and testable. Context provided with minor gaps.
5–6	Question is stated but vague or overly broad. Context partially explained.
0–4	Question unclear or untestable. Little or no context.

2. Data Collection and Sampling — 10 points

Score	Description
9–10	Sampling method described precisely. Potential biases identified and addressed. Sample size justified.
7–8	Sampling method described. Some discussion of limitations.
5–6	Sampling method mentioned but incomplete. Bias not considered.
0–4	Data collection not described or clearly flawed.

3. Exploratory Data Analysis — 15 points

Score	Description
13–15	Appropriate numerical summaries and graphics. Distribution shape, center, spread, and outliers described. Graphics labeled and readable.
10–12	Summaries and graphics appropriate. Some gaps in description.
7–9	Basic summaries provided. Graphics incomplete or missing labels.
0–6	Summaries or graphics missing or inappropriate.

4. Probability Model Selection — 10 points

Score	Description
9–10	Model is appropriate for the data and question. Choice is justified with reasons.
7–8	Model is reasonable. Justification partially stated.
5–6	Model stated without justification.
0–4	Model inappropriate or absent.

5. Inference Procedure — 20 points

Score	Description
18–20	Correct procedure. Assumptions verified. Hypotheses or interval clearly stated. Computation correct.
14–17	Correct procedure. Minor assumption checking gaps. Computation mostly correct.
10–13	Procedure appropriate but assumptions not checked. Some computation errors.
0–9	Wrong procedure, major errors, or inference omitted.

6. Interpretation and Conclusions — 20 points

Score	Description
18–20	Conclusion stated in context. Statistical and practical significance distinguished. Limitations acknowledged.
14–17	Conclusion in context. Practical significance mentioned. Some limitations noted.
10–13	Conclusion stated but not fully contextualized. Significance conflated.
0–9	Conclusion missing, wrong, or purely numerical with no context.

7. Communication: Writing and Graphics — 10 points

Score	Description
9–10	Writing is clear and precise. Technical terms used correctly. Graphics polished, labeled, and purposeful.
7–8	Writing clear with minor errors. Graphics readable.
5–6	Writing adequate but imprecise. Graphics incomplete.
0–4	Writing unclear or technical terms misused. Graphics missing or misleading.

8. Ethics and Reproducibility — 5 points

Score	Description
5	Data sources cited. Limitations of inference stated. Potential for misinterpretation acknowledged.
3–4	Sources cited. Some acknowledgment of limitations.
1–2	Partial citation. No discussion of limitations or ethics.
0	No citation or ethics discussion.

Appendix T

Advanced Statistics and Data Science Readiness Check

Complete these problems before beginning advanced coursework in mathematical statistics, machine learning, econometrics, or quantitative research methods. If you can work through all six sections confidently, you have the probability and statistics foundation required for the next level.

Section 1Probability and Random Variables

Let X have PDF f(x) = 3x² for x ∈ [0, 1]. Find E[X], Var(X), and the median.
Let X ~ Poisson(4). Find P(X ≤ 2) and P(X = 5).
Let X and Y be independent with X ~ Exponential(2) and Y ~ Exponential(3). Find P(X < Y).
Find the MGF of X ~ Binomial(5, 0.4) and use it to find E[X] and E[X²].

Section 2Joint Distributions

Let f(x, y) = 2 for 0 < x < y < 1. Find the marginal distributions of X and Y. Are X and Y independent?
Let X and Y have a discrete joint distribution. Find Cov(X, Y) and ρ(X, Y).
If X₁, …, Xₙ are i.i.d. Normal(μ, σ²), find the distribution of X̄ and of Σ(Xᵢ − X̄)²/σ².

Section 3Estimation and Inference

Find the MLE of p for a Binomial(n, p) model given data x₁, …, xₙ (each 0 or 1).
Construct a 95% confidence interval for μ given x̄ = 42, s = 6, n = 25. Interpret in words.
A study finds a two-sided p-value of 0.03. State what this means and what decision is made at α = 0.05 and at α = 0.01.

Section 4Regression and Modeling

In simple linear regression, β̂₁ = 2.4, SE(β̂₁) = 0.8, n = 20. Test H₀: β₁ = 0 at α = 0.05.
R² = 0.72 and adjusted R² = 0.68 in a multiple regression with 3 predictors and n = 50. Interpret both values.
One-way ANOVA yields F = 6.3 with df = (3, 36). The p-value is 0.001. State the conclusion.

Section 5Bayesian and Likelihood Reasoning

A coin is flipped 20 times and shows 14 heads. Compute the log-likelihood for p ∈ {0.5, 0.6, 0.7, 0.8} and identify which is largest.
Prior: p ~ Beta(2, 2). Data: 8 successes in 10 trials. Find the posterior distribution and compute the posterior mean.

Section 6Stochastic Processes

A Markov chain has transition matrix P = [[0.7, 0.3], [0.4, 0.6]]. Find the stationary distribution.
Events arrive as a Poisson process with rate λ = 3 per hour. Find the probability of exactly 5 arrivals in 2 hours and the expected waiting time until the 4th arrival.

Glossary

Key Terms

Definitions for all major terms introduced in this textbook. Terms are listed alphabetically. Section numbers indicate where each term is first introduced.

alternative hypothesis

The hypothesis contradicting H₀; represents what the researcher suspects may be true. Accepted only when there is sufficient evidence against H₀. (9.3)

Bayes' theorem

P(A|B) = P(B|A)P(A)/P(B); allows updating prior probabilities using observed evidence. (2.11)

Bayesian inference

Framework treating unknown parameters as random variables; updates beliefs through posterior ∝ likelihood × prior. (10.9–10.12)

Bernoulli random variable

Discrete random variable taking values 0 or 1 with probabilities 1−p and p, modeling a single binary trial. (3.8)

beta distribution

Continuous distribution on [0,1] with parameters α, β > 0. Used for proportions and as a conjugate prior for Binomial. (4.12)

bias (estimator)

Bias(T) = E[T] − θ. An unbiased estimator has Bias = 0. (8.4)

binomial distribution

Distribution of the number of successes in n independent Bernoulli(p) trials. (3.9)

central limit theorem

The standardized sample mean converges in distribution to N(0,1) as n → ∞, regardless of population shape (given finite μ and σ²). (7.8)

conditional probability

P(B|A) = P(A ∩ B)/P(A). The probability of B given that A has occurred. (2.7)

confidence interval

An interval [L, U] such that under repeated sampling, (1−α)×100% of such intervals contain the true parameter. The confidence level is a property of the method, not a single interval. (8.7)

conjugate prior

A prior that produces a posterior in the same distributional family when combined with a given likelihood. (10.12)

continuous random variable

A random variable whose probability is described by a PDF; probabilities are areas under the density curve. (4.1)

convergence in probability

Xₙ →^p X if P(|Xₙ−X| > ε) → 0 for every ε > 0. (7.6)

correlation

ρ(X,Y) = Cov(X,Y)/(SD(X)·SD(Y)); standardized covariance measuring linear association on [−1,1]. (5.10)

covariance

Cov(X,Y) = E[XY] − E[X]E[Y]; measures joint variability. Positive = tend to move together. (5.9)

cumulative distribution function

F(x) = P(X ≤ x) for all real x. Non-decreasing, right-continuous, with limits 0 and 1. (3.4, 4.4)

discrete random variable

A random variable whose possible values form a finite or countably infinite set; probability assigned via PMF. (3.1–3.2)

event

A subset of the sample space Ω; a collection of outcomes to which probability can be assigned. (1.4)

expected value

The probability-weighted average of all possible values; the long-run average under repeated sampling. (3.5, 4.6)

exponential distribution

Continuous distribution on [0,∞) modeling waiting times between Poisson process events; characterized by the memoryless property. (4.9)

gamma distribution

Continuous distribution on [0,∞) with shape α and scale β; generalizes the exponential and includes chi-squared as a special case. (4.11)

geometric distribution

Distribution of the number of trials until the first success in i.i.d. Bernoulli(p) trials; discrete memoryless property. (3.10)

hypothesis test

A structured procedure for evaluating evidence against a null hypothesis using a test statistic and p-value. (9.1)

independence (events)

A and B are independent if P(A ∩ B) = P(A)P(B); knowing A occurred gives no information about B. (2.8)

independence (random variables)

X and Y are independent if f(x,y) = fₓ(x)·fᵧ(y) for all (x,y). (5.7)

joint distribution

The probability distribution describing the simultaneous behavior of two or more random variables. (5.2)

law of large numbers

The sample mean X̄ₙ converges in probability (weak) or almost surely (strong) to μ as n → ∞. (7.7)

law of total probability

For a partition {A₁,…,Aₙ}: P(B) = Σᵢ P(B|Aᵢ)P(Aᵢ). (2.10)

likelihood function

L(θ; x) = f(x₁;θ)···f(xₙ;θ); measures how plausible each value of θ is given the observed data. (10.2)

marginal distribution

The distribution of one variable obtained by summing or integrating over the values of the other in a joint distribution. (5.5)

Markov chain

A sequence of random variables satisfying the Markov property: P(future|past, present) = P(future|present). (13.5)

maximum likelihood estimator

The parameter value that maximizes the likelihood function given the observed data. (10.6)

mean squared error

MSE(T) = Var(T) + [Bias(T)]²; combines variability and bias of an estimator. (8.6)

moment generating function

Mₓ(t) = E[eᵗˣ]; encodes all moments and uniquely determines the distribution when it exists in an interval around 0. (6.8)

normal distribution

The bell-shaped distribution N(μ, σ²); the most important in statistics due to the CLT. (4.10)

null hypothesis

The default hypothesis representing no effect, no difference, or the status quo. (9.3)

p-value

The probability, under H₀, of observing a test statistic at least as extreme as computed from the data. (9.5)

Poisson distribution

Distribution of event counts in a fixed interval under a Poisson process with rate λ; PMF = e^(−λ)λˣ/x!. (3.13)

Poisson process

A counting process with independent, stationary increments where counts in intervals of length t follow Poisson(λt). (13.9)

posterior distribution

The conditional distribution of a parameter given the data: π(θ|x) ∝ L(θ;x)·π(θ). (10.11)

power (test)

1 − β = P(Reject H₀ | H₁ true). Power increases with larger n, larger effect size, or larger α. (9.9)

prior distribution

In Bayesian inference, π(θ) expresses beliefs about the parameter before observing data. (10.10)

probability density function

f(x) ≥ 0 with ∫ f(x) dx = 1; P(a ≤ X ≤ b) = ∫ₐᵇ f(x) dx. (4.2)

probability mass function

p(x) assigning probability to each value of a discrete random variable; p(x) ≥ 0 and Σ p(x) = 1. (3.3)

random variable

A function from a sample space to the real numbers, assigning a numerical value to each outcome. (3.1)

random walk

A stochastic process formed by accumulating i.i.d. steps; a simple random walk takes steps of ±1. (13.4)

residual

The difference between an observed value and the fitted value from a regression model: eᵢ = yᵢ − ŷᵢ. (11.7)

sample space

Ω; the set of all possible outcomes of a random experiment. (1.5)

sampling distribution

The distribution of a statistic computed from repeated samples of the same size from the same population. (7.10)

significance level

Pre-specified threshold α below which H₀ is rejected; equal to the Type I error rate under H₀. (9.6)

standard error

The standard deviation of a sampling distribution; SE(X̄) = σ/√n for the sample mean. (7.5)

stationary distribution

π satisfying πP = π and Σπⱼ = 1; the long-run proportion of time a Markov chain spends in each state. (13.8)

stochastic process

A collection of random variables indexed by time or another parameter; models systems that evolve randomly. (13.1)

Type I error

Rejecting H₀ when H₀ is true; its probability equals α. (9.8)

Type II error

Failing to reject H₀ when H₀ is false; its probability is β = 1 − Power. (9.8)

unbiased estimator

An estimator T such that E[T] = θ for all θ; correct on average. (8.4)

uniform distribution

Continuous distribution on [a,b] with density 1/(b−a) everywhere on the interval. (4.8)

variance

Var(X) = E[(X−μ)²]; the expected squared deviation from the mean, measuring spread. (3.6, 4.7)

Answer Key

Selected Answers

Selected answers are provided for odd-numbered problems within chapters. Full solutions are available in the Instructor Edition. For worked solutions to all Appendix Readiness Problems, see the separate Solutions Manual.

A.6.1

∫₀^∞ e^(−2x) dx = [−e^(−2x)/2]₀^∞ = 0 − (−1/2) = 1/2

A.6.2

F′(x) = 3e^(−3x) for x > 0. This is the PDF of an Exponential(3) distribution — the derivative of the CDF yields the PDF.

A.6.4

∫₀¹ x(1 − x) dx = [x²/2 − x³/3]₀¹ = 1/2 − 1/3 = 1/6

A.6.5

Let u = x, dv = e^(−x) dx: ∫₀^∞ xe^(−x) dx = [−xe^(−x)]₀^∞ + ∫₀^∞ e^(−x) dx = 0 + 1 = 1 (This equals Γ(2) = 1! = 1.)

A.6.6

∫₀¹ ∫₀¹ (x + y) dy dx = ∫₀¹ [xy + y²/2]₀¹ dx = ∫₀¹ (x + 1/2) dx = [x²/2 + x/2]₀¹ = 1/2 + 1/2 = 1

T.1 — PDF f(x) = 3x² on [0,1]

E[X] = ∫₀¹ x·3x² dx = 3∫₀¹ x³ dx = 3/4 E[X²] = ∫₀¹ x²·3x² dx = 3∫₀¹ x⁴ dx = 3/5 Var(X) = 3/5 − (3/4)² = 3/5 − 9/16 = 48/80 − 45/80 = 3/80 Median m: ∫₀^m 3x² dx = m³ = 1/2 → m = (1/2)^(1/3) = 2^(−1/3) ≈ 0.794

T.3 — P(X < Y), X ~ Exp(2), Y ~ Exp(3) independent

P(X < Y) = ∫₀^∞ P(X < y) · fᵧ(y) dy = ∫₀^∞ (1 − e^(−2y)) · 3e^(−3y) dy = 1 − 3∫₀^∞ e^(−5y) dy = 1 − 3·(1/5) = 2/5

T.9 — 95% CI for μ: x̄ = 42, s = 6, n = 25

t*(df = 24, 95%) ≈ 2.064 CI: 42 ± 2.064 · (6/5) = 42 ± 2.477 ≈ (39.52, 44.48) Interpretation: We are 95% confident the true population mean lies between 39.52 and 44.48. (The confidence level describes the method's long-run performance, not this specific interval.)

T.14 — Log-likelihoods for coin (14 heads in 20 flips)

ℓ(p) = 14 ln(p) + 6 ln(1−p) (up to additive constants) ℓ(0.5) ≈ 14 ln(0.5) + 6 ln(0.5) = 20 ln(0.5) ≈ −13.86 ℓ(0.6) ≈ 14 ln(0.6) + 6 ln(0.4) ≈ −12.47 ← largest among these four ℓ(0.7) ≈ 14 ln(0.7) + 6 ln(0.3) ≈ −12.79 ℓ(0.8) ≈ 14 ln(0.8) + 6 ln(0.2) ≈ −14.61 Note: The MLE is p̂ = 14/20 = 0.70, which is between 0.6 and 0.8.

T.15 — Posterior: Beta(2,2) prior, 8 successes in 10 trials

Posterior: Beta(2 + 8, 2 + 2) = Beta(10, 4) Posterior mean = 10/(10 + 4) = 10/14 ≈ 0.714

T.16 — Stationary distribution, P = [[0.7, 0.3], [0.4, 0.6]]

Solve πP = π and π₁ + π₂ = 1: 0.3π₁ = 0.4π₂ → π₁ = (4/3)π₂ (4/3)π₂ + π₂ = 1 → π₂ = 3/7, π₁ = 4/7 Stationary distribution: π = (4/7, 3/7)

Index

Section numbers are given. Appendix entries are listed as App A, App B, etc.

A

Adjusted R² 12.8
Alternative hypothesis 9.3
ANOVA 12.10–12.12, App P
Axioms of probability 2.2, App E

B

Bayes' theorem 2.11, 10.9–10.12
Bayesian inference 10.9–10.13, App O
Bernoulli distribution 3.8, App F
Bernoulli process 13.3, App Q
Beta distribution 4.12, App G
Beta function App B.4
Bias (estimator) 8.4, App M
Binomial distribution 3.9, App F

C

Central limit theorem 7.8, App K
Chi-squared distribution App G
Complement rule 2.3, App E
Conditional distribution 5.6
Conditional probability 2.7, App E
Confidence intervals 8.7–8.13, App M
Conjugate prior 10.12, App O
Continuity correction 7.9, App K
Continuous random variable 4.1–4.14
Convergence in probability 7.6, App K
Correlation 5.10, 11.3, App I
Covariance 5.9, App I
CDF 3.4, 4.4

D – E

Degrees of freedom App N
Discrete random variable 3.1–3.14
Ethics (statistical) 14.10, App S
Event 1.4, App D
Expected utility 14.8
Expected value 3.5, 4.6, App H
Exponential distribution 4.9, App G

F – G

F distribution App G
F-test 12.12, App P
Gamma distribution 4.11, App G
Gamma function App B.3
Gaussian integral App B.5
Geometric distribution 3.10, App F

H – I

Hypothesis testing 9.1–9.14, App N
Hypergeometric distribution 3.12, App F
Independence (events) 2.8, App E
Independence (random variables) 5.7, App I
Indicator variables 12.5

J – L

Joint distribution 5.2–5.12, App I
Kolmogorov axioms 2.2, App E
Law of large numbers 7.7, App K
Law of total probability 2.10, App E
Least squares 11.4–11.5, App P
Likelihood function 10.2–10.5, App O
Linear combinations 5.11
Linear regression 11.1–11.14, App P
Log-likelihood 10.5, App O

M – N

Marginal distribution 5.5, App I
Markov chain 13.5–13.8, App Q
MLE 10.6–10.8, App O
Mean squared error 8.6, App M
Memoryless property 4.9, App G
MGF 6.8–6.10, App J
Model assumptions 14.4
Moments 6.7, App H
Multiple regression 12.2–12.9, App P
Negative binomial 3.11, App F
Normal approximation 7.9, App K
Normal distribution 4.10, App G
Null hypothesis 9.3, App N

P – Q

p-Value 9.5, App N
Parameters 1.8, 8.2
Partition App D
Percentiles and quantiles 4.5
Point estimation 8.3, App M
Poisson distribution 3.13, App F
Poisson process 13.9, App Q
Posterior distribution 10.11, App O
Power (test) 9.9, App N
Prior distribution 10.10, App O
PDF 4.2–4.3
PMF 3.3

R – S

R² and adjusted R² 12.7–12.8, App P
Random variable 3.1, App D
Random walk 13.4, App Q
Regression assumptions 11.8, App P
Residuals 11.7, App P
Sample mean 7.3, App L
Sample proportion 7.4, App L
Sample space 1.5, App D
Sample variance App L
Sampling distribution 7.10–7.11, App L
Significance level 9.6, App N
Standard error 7.5, App K
Stationary distribution 13.8, App Q
Stochastic process 13.1–13.14, App Q
Student's t distribution App G

T – V

Transition matrix 13.6, App Q
Tree diagrams 2.12
Type I error 9.8, App N
Type II error 9.8, App N
Uniform distribution 4.8, App G
Unbiased estimator 8.4, App M
Variance 3.6, 4.7, App H

MGU Library Connections

Earlier MGU Texts That Support This Course

Prerequisite Texts

Arithmetic and Pre-Algebra Foundations — Fractions, ratios, proportions, summation notation, and data displays; Chapters 10–14 are directly preparatory
Algebra I and Algebra II — Equations, functions, exponentials, and logarithms; foundational for distribution formulas and transformations
Precalculus and Trigonometry — Polynomial and rational functions, sequences and series, sigma notation; Chapters 9–12 directly relevant
Calculus — Limits, derivatives, and integrals; Chapters 5–9 are the calculus foundation required for continuous random variable material in Chapters 4–7 of this text
Discrete Mathematics — Combinatorics, sets, logic, and graph theory; Chapters 6–9 directly support Chapter 2 (probability rules) and Chapter 3 (discrete distributions)
Linear Algebra — Matrix operations, systems, and vector spaces; Chapter 4 supports the regression material in Chapters 11–12
Differential Equations — Supports advanced continuous modeling and stochastic differential equations in post-course study

Concurrent Courses That Connect Well

Concurrent Study

High School Statistics (MGU) — A conceptual introduction; this calculus-based text provides the mathematical foundation beneath those ideas
Calculus II — Sequences, series, and additional integration techniques directly support MGF theory and limit theorem proofs
Calculus III — Multivariable calculus and double integrals are used in Chapter 5 joint distribution theory

Later MGU Texts This Course Supports

Paths Forward

Calculus II and Calculus III — Can be studied alongside or after this text; multivariable calculus deepens joint distribution theory
Numerical Methods — Simulation, root-finding for MLEs, and numerical integration connect directly
Data Science Foundations — Applies the probability and statistical inference framework to data pipelines, feature engineering, and exploratory analysis
Machine Learning Mathematics — Probabilistic models, MLE, Bayesian networks, expectation-maximization, and risk minimization extend the core ideas of Chapters 7–10 and 14

Reference Materials in the MGU Library

MGU Reference Collection

MGU Mathematics Dictionary — Entries for all key terms in the Glossary of this text
MGU Formula Sheet Collection — Probability, distributions, inference, and regression formulas; corresponds to Appendices E through P
Distribution Tables — Standard normal (Φ table), t critical values, chi-squared critical values, F critical values; available in print and digital formats
MGU Statistical Computing Guide — Extended R and Python tutorials for every procedure in this textbook
MGU Capstone Project Archive — Annotated example projects at Distinguished, Proficient, and Developing levels for each type of analysis

External Resources for Further Study

Recommended Advanced Texts

Mathematical Statistics with Applications — covers mathematical statistics with full proofs

Bayesian Data Analysis — extends the Bayesian foundations of Chapter 10

An Introduction to Statistical Learning — applies regression and probabilistic modeling foundations to machine learning

Stochastic Processes (Sheldon Ross) — extends Chapter 13 into full Markov chain theory, Poisson processes, and Brownian motion

The American Statistical Association (ASA) — professional organization with student resources, ethical guidelines for statistical practice, and statements on reproducibility

Series Placement

Probability and Statistics in the MGU Pathway

This text follows Calculus in the MGU sequence and may be studied alongside Linear Algebra, Differential Equations, or Discrete Mathematics depending on the student's pathway. It serves as the bridge from calculus and algebraic modeling into mathematical statistics, data science, machine learning, econometrics, quantitative finance, engineering reliability, scientific research, public policy, medicine, and advanced decision-making under uncertainty.

Back MatterAppendices, Glossary,Answer Key & Index