R Programming Leveling Course

Short, intensive R leveling course for students entering the Master's programs in Economics and in Management Sciences (Econometría I). Covers R and RStudio fundamentals, reproducible project workflows, data wrangling with the tidyverse, visualization with ggplot2 and descriptive statistics on real GEIH microdata, and basic regressions (Mincer equation with lm, fixest, and modelsummary) — with a critical, verified use of generative AI built into every unit.

Instructor: Eduard F. Martínez-González

Institution: Universidad ICESI — CIENFI — Centro de Investigaciones en Economía y Finanzas

Program: Master's in Economics and in Management Sciences

Term: July 2026

Original title (in Spanish): Curso Nivelatorio de Programación en R — Manejo y análisis de datos para investigación aplicada.

All course materials — theory documents with runnable R in the browser (via webR), guided practices, assignments, quizzes, AI activities, and the course datasets — are published on the course website:

  Course website   Full program (syllabus)

Course description

A brief, intensive leveling course so that incoming Master’s students arrive with a common base in R: importing, cleaning, transforming, merging, describing, and visualizing data, estimating basic regressions, and — above all — working in an organized and reproducible way, as applied research demands. No prior programming experience is required.

The course runs on a see → replicate → apply cycle: students first watch the theory (documents embed executable R code in the browser, so no installation is needed to start experimenting) and the support video, then replicate the guided practice in RStudio inside their own project, and finally apply what they learned in an assignment with its quiz and an AI activity. Synchronous sessions open each block; most of the learning happens by doing.

The four units

  1. Fundamentals of R and a reproducible workflow. What R does when code runs, how to read errors and warnings, vectors and data frames, and the foundation of the whole course: RStudio projects, organized folders, and well-styled scripts.
  2. Data wrangling with the tidyverse. The 80% of real empirical work: importing and diagnosing, cleaning a dirty dataset with documented decisions, transforming with dplyr, grouping, and merging datasets with joins.
  3. Visualization and descriptive statistics. From clean data to a communicable finding with ggplot2 and descriptive tables, working on a real extract of Colombia’s GEIH household survey (DANE).
  4. Basic regressions and the empirical workflow. The Mincer equation on the GEIH: lm, interpretation in units (and without unjustified causal claims), fixest, tables with modelsummary, and the full reproducible pipeline.

Course materials

Unit Theory Video Practice Assignment Quiz AI activity
1. Fundamentals Theory Video Practice 1 Assignment 1 Quiz 1 Activity 1
2. Data wrangling Theory Video Practice 2 Assignment 2 Quiz 2 Activity 2
3. Visualization Theory Video Practice 3 Assignment 3 Quiz 3 Activity 3
4. Regressions Theory Practice 4 Assignment 4 Quiz 4 Activity 4

Evaluation

Component Weight Allowed use of AI
Assignments 1–4 (script + quiz) 40% To understand and debug, declared
AI activities 10% Designed to use AI with verification
Final project (reproducible pipeline + report) 35% Integrated, with a usage appendix
Oral defense of the code 15% No AI

The AI policy in one sentence: your brain first, then the AI; everything the AI produces gets verified; and every use is declared. Levels and the verification checklist are in the AI usage guide.

Final project

A complete empirical analysis, in pairs, on the firm innovation dataset: clean with documented decisions, build sector-level indicators and compare them against the national benchmark, plot, estimate the innovation–sales relationship and interpret it — all inside a reproducible project defended orally. Project statement · Rubric

Datasets

Dataset Unit of observation Used in
innovacion_empresas.csv Firm (506, deliberately dirty) Units 1–2 and final project
sectores_agregado.csv Sector (7, clean benchmark) Unit 2 and final project
geih_nivelacion.csv Employed person (21,821, real GEIH) Units 3–4

Variable-by-variable description: data dictionary.

Course guides

Calendar

Four synchronous sessions (10:00–12:00) in July 2026 — July 14, 21, 27, and 31 — each opening a block of asynchronous work: theory and practice, assignments, AI activities, and the final project in pairs, which is presented in the last session.