Statistics — Complete Study Notes

Complete study notes from the Inferential Statistics course sessions — covering descriptive statistics end to end, from what a variable is through to building and reading a box plot, with every worked example, Excel walkthrough and industry application from class.

Everything in this domain is tagged inferential-statistics, so the whole set is findable as one group.

A note on the naming

The course session is called “Inferential Statistics”, but the material in these notes is Chapter 1 — Descriptive Statistics. That’s the course’s own chapter mapping:

Chapter	Topic	Covered here
Chapter 1	Descriptive Statistics	Yes — all of these notes
Chapter 2 onwards	Inferential Statistics	Deferred to later sessions

So: the session name is Inferential Statistics; the content is descriptive statistics — the foundation the inferential material is built on. The distinction between the two is itself the first topic in the notes.

Source reliability key

The notes carry a tag on claims so you know how well-supported each one is:

Tag	Meaning
[T]	Supported by class transcript — the instructor’s spoken explanation was captured
[S]	Reconstructed from slide screenshots only — the primary session’s recording stopped at ~1:14:00 while the class ran to 1:51
[T+S]	Both transcript and slides available

Anything tagged [S] has accurate slide content but no verbal commentary behind it.

Session details

• Course: Unit 1 — Descriptive Statistics
• Primary session: “Inferential Statistics” · 9 August 2026 · 1 h 51 m · 952 transcript entries · 30 screens
• Also drawn from: the variance/standard-deviation session (Excel exercises), the quartiles/IQR/box-plot session, and the box-plot practice session

The notes, in order

Part I — Foundations

Descriptive vs. inferential statistics and how to tell them apart, the five opening classification examples worked through, and why statistics matters in practice: A/B testing (the control/treatment framework), feature engineering, and AI model bias detection via the Equal Opportunity Score.

Part II — Variable Types

The full taxonomy — numerical (discrete, continuous) vs categorical (nominal, ordinal), plus unstructured data. Includes the two classification traps: PIN codes that look discrete but aren’t, and encoded categories that look numerical but aren’t.

Part III — Central Tendency

Mean, median and mode: formulas, merits and demerits for each, the rules for filling missing values, what outliers damage and what survives them, and how skewness pulls the three measures apart.

Part IV — Dispersion

Why an average alone never tells the whole story. Range (and how one value flips a verdict), variance, standard deviation and why it exists, the population vs. sample n − 1 question, two Excel walkthroughs, and the coefficient of variation for comparing consistency across different units.

Part V — Quartiles, IQR & Box Plots

Quartiles and the interquartile range, the 1.5 × IQR outlier rule, the five-number summary, and how to construct and interpret a box plot — with two fully worked examples and the restaurant wait-times practice problem.

Part VI — Applying It

Which measure to use when, outliers as anomaly detection vs. noise (with the IT job-completion-time case), standard deviation for feature selection, and where statistics sits in an AI/ML workflow.

Reference

The complete formula sheet, every number used in class, the exam traps, the student questions answered in class, what’s deferred to later chapters, and the ten things worth memorising.

The three sentences that carry the whole session

1. Descriptive statistics describes the data you have; inferential statistics generalises to data you don’t.
2. Central tendency tells you where the data sits — dispersion tells you how much you can trust that.
3. An outlier is either the error you must remove or the signal you were looking for; statistics tells you it’s there, context tells you which.