ECON 371 Class Recording 8/25

Overview

Professor Lantis introduces ECON 371’s course structure and its goal of estimating causal relationships rather than mistaking correlation for causation. He explains the grading and study expectations, reviews data types and descriptive statistics, demonstrates probability-weighted calculations with IU football data in Excel, and prepares students to use Stata through IUanywhere, Citrix, and OneDrive.

Key takeaways

ECON 371’s grade consists of a 30% midterm, 30% final, six Canvas quizzes worth 12% total, six Stata problem sets worth 24%, and in-class quizzes worth 4%; the four lowest in-class quiz scores are dropped.
The course teaches causal inference by showing why correlation can mislead: high-smoking areas may receive higher cigarette taxes, and well-funded districts may have both smaller classes and higher test scores for other reasons.
Cross-sectional data cover multiple entities at one time, time-series data track one entity across time, and panel data combine both dimensions, as in countries’ GDP recorded over multiple years.
For a discrete random variable, the mean is the probability-weighted sum of possible outcomes, and variance is the probability-weighted sum of squared deviations from that mean.
Stata is accessed through IUanywhere and may require Citrix Workspace; students must authorize IU cloud storage and keep course datasets in OneDrive for Stata to open them.
Course deadlines are strict: late work receives zero, while partial submissions can earn credit; starting quizzes and Stata problem sets early also leaves time to seek help.

Chapters

0:00 ECON 371 Course Resources, Recording, and Classroom Expectations

Professor Lantis says Canvas slides and class material are the primary course resources; the assigned textbook is optional background reading.
Class recordings will be posted to a course-specific YouTube playlist, but iPad notes may not appear in the recording.
Students will need Canvas access for quizzes and are strongly encouraged to bring a laptop for Stata work.
Midterm and final exams require a physical calculator with basic arithmetic and powers; phones are not allowed.

7:29 ECON 371 Grade Weights and Canvas Quiz Schedule

The midterm and final each count for 30% and include substantial written-response work that requires showing calculations.
Six Canvas quizzes count for 12% total; each covers roughly two weeks of material and is posted after Tuesday class.
Canvas quizzes are due the following Monday, allow 60 minutes once started, and should be started by 10:59 p.m. to meet the deadline.
Practice questions are provided in advance, and quiz answers become visible after the deadline.

12:05 Comprehensive Exams and Attendance-Based In-Class Quizzes

Professor Lantis describes exams as comprehensive rather than fully cumulative: later topics build on earlier material, though the final focuses on the second half.
In-class Canvas quizzes are brief, usually one or two questions with 5–10 minutes to respond, and contribute 4% of the course grade.
Students can expect roughly 16–20 in-class quizzes, with the four lowest scores dropped to accommodate illness or personal absences.
Attendance is not formally required, but quizzes occur unpredictably and missed classes make it harder to keep up with the course’s cumulative concepts.

16:00 Six Stata Problem Sets and What Exams Test

Six problem sets count for 24% total, or 4% each, and generally provide about 1.5–2 weeks to complete.
Assignments use Stata data files to create variables, run regressions, and answer questions about data; Thursday is the typical deadline.
Students can rely on in-class code and examples as templates for completing each problem set.
Exams do not test Stata syntax; they provide Stata output and ask students to interpret results and answer related questions.

19:20 Course Deadlines, Study Practices, and Office Hours

Professor Lantis advises checking exam dates early and arranging an exam conflict at least a week in advance; make-up arrangements typically require taking the exam early.
Reviewing slides before class, attending consistently, and working through practice problems and practice exams are the recommended study routine.
Missed deadlines receive zero credit rather than a gradual late penalty, so submitting partial work is better than leaving a quiz or problem set blank.
Professor Lantis holds office hours in room 309 from 8–11 a.m. Monday and Wednesday, with appointments available when that schedule conflicts; the GA’s hours were still being arranged.

28:00 Econometrics’ Central Goal: Estimating Causal Effects

The course aims to estimate causal relationships, not merely describe correlations between economic variables.
Cigarette taxes may correlate with smoking because governments could tax places where smoking is already common; that correlation alone does not show taxes increase smoking.
Class size and test scores may be correlated because well-resourced districts can both hire more teachers and provide other advantages that raise scores.
Applications include testing whether race affects mortgage denials or rates and whether cigarette taxes actually reduce smoking.

33:00 Cross-Sectional, Time-Series, and Panel Data

Cross-sectional data record multiple entities at one time, such as every country’s GDP in 2024 or students’ current GPAs.
Time-series data follow one entity across multiple periods, such as U.S. unemployment rates over time or one country’s GDP across 50 years.
Panel data combine multiple entities and multiple periods, such as student test scores tracked across grades or countries’ GDP across years.
An entity can be a person, city, country, or company; panel data combine features of cross-sectional and time-series data.

38:20 Population, Samples, Random Variables, and Distributions

Uppercase N denotes population size: population data would include every unit of interest, while a survey of 100 or 1,000 people is a sample.
A random variable represents an outcome that can take different values when an observation is selected, such as a school’s average test score.
A distribution describes possible outcomes and their associated probabilities or frequencies.
For a discrete variable, one can calculate the probability of a specific value; for a truly continuous variable, the probability of one exact value is zero, so probabilities are defined over ranges.

44:00 Summation Notation, Mean, Median, and Mode

A summation from 1 to 5 adds the indexed values across five observations; for example, it can total the number of courses taken by students in a dataset.
For observed data, the mean is the sum of values divided by the number of observations; for a random variable, it is the sum of each possible value multiplied by its probability.
The median divides ordered observations in half; with an even number of observations, average the two middle values.
The mode is the most frequent observed value, or the value with the highest probability for a discrete random variable.

51:30 Variance and Standard Deviation for Data and Random Variables

Variance averages squared deviations from the mean, emphasizing observations that lie farther from the center.
Standard deviation is the square root of variance, returning dispersion to the variable’s original scale.
For a random variable, each squared deviation is weighted by the probability of its corresponding outcome.
Professor Lantis previews using variance and standard deviation in later test-statistic and confidence-interval calculations.

54:57 Excel Setup and IU Football Win-Count Data

Professor Lantis uses an Excel dataset of IU football seasons from the previous 25 years to illustrate discrete random-variable calculations.
The dataset lists possible win totals and the frequency of seasons with each total; zero wins are omitted because none occurred in the sample.
To convert a frequency into a probability, divide it by the 25 seasons—for example, six seasons with a given win total produce probability 6/25.
Excel dollar signs lock a cell reference when formulas are copied down, allowing a shared denominator or mean to remain fixed.

59:30 Calculating IU Football’s Expected Wins and Variance

Expected wins are calculated by multiplying each possible win total by its probability and summing the products; the example yields a mean of 4.56 wins.
To calculate variance, subtract 4.56 from each possible win total, square the deviation, and multiply by that outcome’s probability.
Summing the probability-weighted squared deviations gives variance; taking its square root gives the standard deviation.
Ignoring probabilities would incorrectly assign equal weight to all win totals from 1 through 12, producing 6.5 even though some outcomes occur much more often than others.

1:04:10 Setting Up Stata Through IUanywhere, Citrix, and OneDrive

Students access Stata through IUanywhere using IU credentials; some devices may require installing and enabling Citrix Workspace.
Professor Lantis points students to a setup walkthrough on his YouTube playlist and recommends using a laptop because the tablet interface is limited.
Authorize IU cloud storage so Stata can access files in the student’s OneDrive, then create an ECON 371 folder for course datasets.
Before Thursday, students should open Stata and try loading a Canvas data file from OneDrive to confirm the setup works.

1:11:00 Keeping Pace with ECON 371 and Preparing for Thursday

Professor Lantis warns that the course moves quickly and that econometrics and Stata may initially feel unfamiliar, even during review material.
Starting problem sets early leaves time to troubleshoot Stata and ask questions during office hours instead of getting stuck near a deadline.
He encourages students to use office hours, complete practice problems, and balance studying with other parts of college life.
The next class will begin hands-on Stata work with data; Professor Lantis also offered to be available from 8–11 the following morning.

Summary, takeaways, and chapters were generated by AI from the video's transcript and may contain errors. The video belongs to its creator, Professor Lantis.