EC 285 - Lecture 01 - 2026-09-10

Overview

First lecture of EC 285 — course introduction and outline, with essentially no technical content beyond framing. Liaqat set up the central idea of the course in one line: how we can use a sample of individuals to make well-reasoned claims about an unobserved population, and unpacked it into the sample/population distinction, the role of probability and uncertainty, and hypothesis testing as the eventual destination (Chapter 9). She then split the course into its two halves — descriptive statistics (charts, summaries, spotting outliers) and the much larger inferential half — and drew out the paired terms she'll keep testing: descriptive vs. inferential, parameter vs. statistic, correlation vs. causation. The rest of the session was administration: grading scheme, tutorials, Stata assignments, the AI policy, the custom AI Tutor, and a heavy pitch on why attending lectures and taking your own notes matters.

Key concepts

• Statistics (what the course is) — the study of tools for processing, summarising and analysing data so you can make decisions and claims about a larger population you can't observe. Her one-line summary, which she asked the class to reread: "how we can use a sample of individuals to make well-reasoned claims about an unobserved population."
• Population vs. sample — the population is everyone (all Canadians, all Laurier students); the sample is the subset you actually survey. You sample because surveying everyone is slower, more expensive and effectively impossible — "even Statistics Canada does not go and collect data for every single Canadian." Notation: sample size is lowercase n, population size is uppercase N.
• Representativeness and sample size — a sample only supports good predictions "if you do it properly and if your sample is representative enough"; the larger n, the better the results.
• Uncertainty and probability — probability is the likelihood of an outcome, a number between 0 and 1 ("if it rains tomorrow is 30%"). We make decisions under incomplete information every day, so conclusions from a sample are never 100% certain; probability distributions are how we attach as accurate a number as possible to an event.
• Statistical significance — the credibility attachment on a result. Her framing: a study is publishable/reliable when the results make sense, have a story behind them, and carry a high enough probability. "If it's only a 50% probability that this will be true for the larger population, then it's not worth anybody's attention."
• Hypothesis testing — you hold a hypothesis and test it on a representative sample rather than the whole population. This is Chapter 9, and "there is a lot of work that we have to do before we get to" it.
• Descriptive statistics — the first half of the course: turning an unreadable spreadsheet into pie charts, bar charts and histograms so patterns become visible. Also how you spot outliers — an unusual point may be a genuine extreme or a data-entry error worth going back to check.
• Effect of outliers — in a large class a couple of outliers barely move the average, but "if it's a small sample, then including outliers can really bias your averages and median and standard deviation." Standard practice: report the mean with and without them.
• Inferential statistics — the second and much larger half (most chapters): forecasting, prediction, hypothesis testing.
• Parameter vs. statistic — a parameter describes the population; a statistic describes the sample (a sample's average grade, average age). "If I'm saying the parameter is equal to 50, it means I'm referring to the population."
• Correlation vs. causation — correlation is two variables moving together; causation is X producing Y. Her example: you eat ice cream at the beach and get sunburn an hour later — ice cream and sunburn are correlated because both track hot weather, but the heat (and no sunblock) causes the burn, not the ice cream.
• Where this course sits — it is not a typical economics course: "there's no economics, it's all math and formula." No supply-and-demand diagrams or PPFs; regression and real data analysis are EC 295 and EC 395, which she does not teach.

Flagged for exams

• Descriptive vs. inferential statistics — stated outright: "This, by the way, is a potential multiple choice question on the midterm, where I ask you the difference between descriptive and inferential statistics."
• Parameter vs. statistic — "another potential multiple choice question." She stressed these are "tiny differences in single terms that we will be using, I'll be calling them out, so you should know what the difference is."
• Correlation vs. causation — "that's the distinction and it's important and we talk about this quite a few times actually in the course."
• Midterm 1 covers Chapters 1, 2 and 3 only — deliberately less material than in earlier terms because the date is early.
• Exam format — all exams are a mix of conceptual multiple-choice and numerical/math problems. "For now it's 50/50, but it could be 60/40, or 40/60." You choose how to split the 60 minutes between the two parts; there is no separate timing per section.
• Formula sheet and aids — a formula sheet and statistical tables are provided, and she will show you the formula sheet in advance: "what's not on the formula sheet obviously you have to remember." Calculators allowed; no cheat sheet or other memory aid.
• Best exam practice is tutorial questions, the examples she works in class, and end-of-chapter textbook questions. Assignments are Stata-based and are explicitly not good sample exam questions.
• Class-only material — repeated several times: the posted slides are "deliberately very brief… just an outline," not comprehensive notes. Extra practice problems, extra formulae, and "the hints that I give away about what is it that you should focus on when you are studying for exams" only happen in the room. "That's why you need to come to class."
• Her exam style — "My exams, there's no surprises… It's easy if you're prepared." For a 200-level course she doesn't consider them difficult if you study. She will take in-class practice problems and "change a little bit on the exam."
• Notation to have straight: sample n lowercase, population N uppercase.
• Some lecture material is not in the slides and some is not in the textbook — read the book and your own notes together when preparing.

Admin & deadlines

• Homework due before next class (Tuesday, 4:00 PM): read the full 7-page course outline — "there is some time sensitive things in there." Reading Chapter 1 of the textbook is encouraged but optional this week; the course begins properly on Tuesday, picking up the remaining Chapter 1 material she didn't reach.
• Assessment (syllabus): Midterm 1 20%, Midterm 2 20%, two assignments @ 7.5% = 15%, tutorial participation 10%, final exam 35% — or 55% if the lower midterm is dropped.
• Midterm 1: Sunday, October 4, 12:00–1:00 PM (syllabus time; in class she said only "October 4th" and "less than a month away"). One hour, combined sitting for all three sections. Covers Chapters 1–3. Her explanation of why that date was chosen is garbled on the recording — the gist is a room-capacity constraint forced a combined weekend slot, and she wanted it done before reading week.
• Midterm 2: Saturday, November 21, 3:00–4:00 PM (syllabus; not mentioned by date in class).
• Final exam: cumulative, during the exam period, 2 hours. She noted the length "might change… might go up a little bit, not down."
• Dropping a midterm is automatic. Asked twice, she confirmed: at the end of term she computes all three schemes — (1) all components as weighted, (2) drop Midterm 1, (3) drop Midterm 2 — and takes the highest. "You don't have to decide anything… please don't email me." (Separately, the syllabus's missed-midterm rule still applies: a missed midterm is a zero unless an Illness Self-Declaration form is filed within 2 days, in which case the weight shifts to the final.)
• Assignment 1 (Stata): due October 23; Dropbox closes October 26. Late penalty 10% per day for up to 3 days, then the Dropbox closes. Assignments and due dates are already posted on MyLS.
• Assignments are meant to be easy points — a sample assignment with worked answers is already in the Assignments folder; you change the data and variable names and run the same code. Marked leniently.
• AI policy: generative AI is permitted, and applies only to assignments (everything else is paper-based and in person). You must cite its usage comprehensively — idea generation, writing, fixing or verifying code. She expects you to understand the code you submit. Her framing: use it to learn, not as a shortcut — "you may get a good grade, but in the long term, it's your loss."
• AI Tutor: an optional custom GPT built for this course, linked from the course outline; needs a ChatGPT account. It knows the OpenStax textbook, the course outline and some of her lecture notes — it has no access to MyLS, grades or student data. Configured not to hand you the answer immediately. It's an experiment ("I was testing this out last night"); ChatGPT makes mistakes, so try problems yourself first and tell her if answers look wrong.
• Textbook: OpenStax Introductory Business Statistics 2e, required, free and downloadable — "you have no excuses not to do your reading." Follow the weekly reading schedule in the outline (Week 1 = Chapter 1). She noted the optional second text is expensive and referred to it as "Newbold's"; the syllabus lists the optional text as OpenIntro Statistics, 4th Ed. 2019.
• Tutorials (10%): no tutorials this week (Thursday/Friday of Week 1) — they start next week. Attend the tutorial you are registered in; no exceptions, because your IA marks your section's work. Each tutorial has a short easy-to-medium problem submitted on paper (hard copy only): 1 mark if correct, 0.5 for submitting — even just your name gets you the participation half — 0 if you don't show. Two lowest tutorial grades are dropped, so you can miss two. Tutorials also deliver the Stata labs and extra practice problems she won't have time for in lecture. No tutorials during midterm exam weeks.
• Stata labs are posted on MyLS under Content, along with all slides for the term.
• Lecture attendance: not marked (she dropped iClickers after integrity problems last term), but "it is absolutely essential to attend lectures if you want to do well." Bring a notebook and work the in-class problems.
• Section attendance: you must attend the section you're enrolled in — all sections are full and it's a fire/safety regulation. ~350 students across A, B and C. One-off exceptions (e.g. a job interview) need an email to her in advance; not a routine.
• Room: Section B is in SB106, Tuesday & Thursday 4:00–5:20 PM. She dislikes the single board here and has requested a move to Lazaridis Hall, but expects it won't happen. She'll try to use the lower boards so everyone can see.
• Contact: virtual (Zoom) office hours Monday 12:00–2:00 PM, first-come first-served from the waiting room; other times by appointment. Office LH 2019. Email replies within 24–48 hours excluding weekends. Put your section number (EC285A/B/C) in the subject line with your name and student ID. Don't email about things answered in the outline. Support available: 3 TAs, 3 IAs, the instructor, and the AI Tutor.

Transcribed automatically from the lecture recording. Course info used: EC 285 - Syllabus - Fall 2026.pdf. Audio archived at /mnt/porsche/configs/lectures/archive/2026-09/EC 285 - 2026-09-10 - Lecture 01.opus.