Preface

Why do we learn Mathematical Statistics?

Why study mathematics when software can fit models and AI can generate answers? Mathematical statistics helps us understand what those answers mean, when to trust a method, and how to approach unfamiliar problems. Its value lies in both the results and the habits of thought you develop. Learning a proof is also practice deciding which details matter and explaining why a conclusion follows.

• Sharpen your mind. Precise definitions, logical arguments, and counterexamples help you distinguish what is known, assumed, or still unproved. Struggling with a proof can reveal exactly where your understanding needs work.

• See the benefits of abstraction. A common language connects seemingly different problems. Sums, areas, and probability-weighted averages can all be studied through integration: one principle helps explain several kinds of calculation.

• Think rigorously about data. Statistical claims depend on models, assumptions, and measures of success. Mathematics helps you distinguish an observed pattern from a justified conclusion and understand the limits of a guarantee.

• Build understanding for research. Knowing why a method works lets you ask which assumptions could be weakened or how an argument might extend. Such questions help you read papers critically and develop ideas.

• Learn deeply and clearly. Connecting definitions, examples, pictures, and proofs makes ideas easier to recover and use. Aim for understanding you can explain after a lecture’s details have faded.

The aim is independent, discerning thinking about uncertainty, evidence, and statistical methods. You do not need to love every technical detail.

How to use these notes

These notes guide your study of STAT 709 through definitions, examples, calculations, and proofs. Use them alongside a textbook and your own problem solving. Each lecture has learning goals, a lecture map, priority labels, and exercises. Start with high-priority material, then work through mid-priority connections. Explore optional material as your interests and time allow.

Walking through the webpages

Students bring different backgrounds in probability, analysis, linear algebra, and statistics. Learn at your own pace and spend your study time deliberately. You need not open every block in one sitting.

Choose a reading view.

Core only shows high-priority material for a first pass. High + mid, the default, covers examined reading, subject to the lecture-specific exemptions and optional Part VI. Show all includes optional material. A core-only pass is not complete exam preparation.

Use help selectively.

Quick Recall reviews earlier concepts; Quick Explanation explains incidental terminology; Instructor’s Note adds commentary. Follow their links when you need the full context.

Use figures actively.

Predict what will happen before moving a slider or changing an example. Connect the result to a definition or theorem. Figures guide intuition but do not prove general claims.

Choose practice independently.

Core practice includes high-priority exercises; Exam practice, the default, includes high and mid; All practice adds enrichment. Reading views and practice filters operate independently. Solutions stay closed until you choose to open them.

Balance time and benefit. Give high- and mid-priority material regular attention. When stuck, identify the main claim and the step blocking your understanding, then return with a focused question. Learn as much as you can within your available time. Judge progress by what you can explain and use, rather than by the number of pages you click through.

You need not understand every optional extension or linked paper. If interested, explore a few advanced ideas deeply instead of skimming everything. Let that exploration build on your command of the required material.

Balance rigor and intuition

The notes combine intuitive explanations and figures with rigorous statements and proofs. Intuition suggests what might be true; proof establishes it under precise assumptions. Practice translating between the two.

1. Start with an idea or picture. Describe the behavior you expect in ordinary language or a simple example.

2. Make the claim precise. Name the objects, assumptions, and conclusion. What exactly does “small” or “converges” mean?

3. Justify the steps. Identify the definition, inequality, or theorem behind each move and check its hypotheses. Separate the main idea of an argument from the algebra that implements it.

4. Translate back. Interpret the result, try a boundary case, and ask what fails without an assumption.

A picture may suggest that curves settle down. Before moving a limit through an integral, ask what controls their areas. Convergence theorems turn that intuitive question into conditions you can check and use in a proof.

Self-study is expected!

Read all high-priority and mid-priority topics on your own, including material not covered line by line in class. Both tiers are examined except for the optional Part VI and explicitly stated lecture-specific exemptions. Bring unresolved questions to class or office hours.

Why the exercises matter

Exercises make definitions usable and expose gaps in understanding. Try several times before consulting books, classmates, AI, or the provided solution. Draw a picture, check a simpler case, unpack a definition, or try another theorem. Main-text prompts link to complete exercises and solutions in the Practice section.

After seeking help, close the explanation and reconstruct the argument in your own words. Identify the missing move and why it works. Use complete solutions for review after making serious attempts. Try explaining the same idea again without looking at the symbols. A complete solution is a model of an argument, not a replacement for producing your own.

Externalize your thoughts

Write, sketch, or speak your explanation. Self-explanation makes your reasons and uncertainties visible. Ask:

• What am I trying to show, and what do I know?

• Why choose this definition, example, or theorem?

• Which assumption justifies this step? What remains?

• Can I explain the conclusion in ordinary words?

An analogy is chain-of-thought prompting in large language models: intermediate reasoning examples can help a model solve a task, as studied by Wei and colleagues. Borrow the habit of making steps explicit and inspectable, without assuming human learning works like a model. Always check the mathematics, even when an explanation sounds fluent.

Volunteering for a mini-presentation

Volunteer to explain one topic in about 10 minutes: an example, a proof step, a connection, or a question you have explored. Contact me to volunteer or discuss a topic. A possible outline is:

1. Two minutes: set up the question. Introduce the definitions and assumptions your audience needs.

2. Five minutes: explain the argument. Externalize your thoughts. Tell classmates why you choose each theorem, example, bound, or rearrangement.

3. Two minutes: check and interpret. Try a small example, identify a limitation, or explain why an assumption matters.

4. One minute: give the takeaway and invite a question. Leave the class with one idea they can explain or use.

Aim for clarity and shared learning. You need not cover an entire lecture or give a perfect performance. Explaining where you were stuck and what helped can be especially useful.

Raise questions

Questions are welcome in class. Ask about unclear definitions, assumptions, or derivation steps. Your question need not be polished; others may share the same uncertainty.

Finding typos, inconsistencies, missing assumptions, or confusing explanations is valuable. Include the lecture and section, quote the relevant statement, and describe your difficulty. For webpage issues, include a link. A proposed correction is welcome but not required.

Choose the channel that fits your question:

• Report an issue: errors or unclear explanations in the notes, figures, or website. The link carries the page and section; choose a public GitHub report or a private email. Public reports must not include homework solutions or personal information.

• Canvas Discussions: questions about the course material. You can also ask in class or during office hours.

• Email: private matters.

• Comments (Disqus): public conversation about ideas in each lecture and this preface. Use Report an issue for corrections so they can be tracked through a fix.