Part II. Exploratory Data Analysis and Visualization

NREC4107 - Applied Econometrics

Purpose of Part II

Part II introduces exploratory data analysis and visualization. Before estimating a regression model, an applied researcher should understand the data.

Graphs help us see distributions, unusual observations, group differences, and possible relationships among variables. They do not prove an economic argument by themselves, but they help us ask better empirical questions.

In this part, we continue using the milk product dataset as the main running example. The dataset includes variables such as Price, Size, Pieces, Volume, Brand, Type, Fat, Fresh, Package, Flavor, and Location.

Why visualization matters

A table gives precise numbers. A graph gives structure.

A good graph can show:

whether prices are concentrated or widely dispersed
whether some brands are more expensive on average
whether package size is associated with price
whether categorical variables such as fat content, package type, or location matter
whether a regression question is worth asking

A poor graph can also mislead. For this reason, every graph in this course should be interpreted carefully.

Chapters in Part II

Chapter	Topic
Chapter 5	Descriptive Statistics for Food and Agricultural Data
Chapter 6	Univariate Graphs
Chapter 7	Bivariate Graphs
Chapter 8	Multivariate and Interactive Graphs
Chapter 9	From Graphs to Research Questions

Dataset and software

All code examples are written for Google Colab. Students should place the dataset in this folder structure:

MyDrive/
  NREC4107/
    data/
      Milk_Data_Cleaned.csv
    notebooks/
    outputs/

The code shown on the website is for learning and should be run in Google Colab, not on the website.

What students should be able to do after Part II

After completing Part II, students should be able to:

describe numeric and categorical variables
create basic summary tables
produce univariate, bivariate, and multivariate graphs
identify unusual values and possible data problems
distinguish visual patterns from causal claims
convert visual observations into applied econometric questions

Key message

Exploratory data analysis is not a decorative step. It is part of empirical reasoning. A good graph should help us understand the data and prepare for careful modeling.

--- title: "Part II. Exploratory Data Analysis and Visualization" subtitle: "NREC4107 - Applied Econometrics" format: html: toc: true code-fold: false code-tools: true execute: eval: false warning: false message: false --- ## Purpose of Part II Part II introduces exploratory data analysis and visualization. Before estimating a regression model, an applied researcher should understand the data. Graphs help us see distributions, unusual observations, group differences, and possible relationships among variables. They do not prove an economic argument by themselves, but they help us ask better empirical questions. In this part, we continue using the milk product dataset as the main running example. The dataset includes variables such as `Price`, `Size`, `Pieces`, `Volume`, `Brand`, `Type`, `Fat`, `Fresh`, `Package`, `Flavor`, and `Location`. ## Why visualization matters A table gives precise numbers. A graph gives structure. A good graph can show: - whether prices are concentrated or widely dispersed - whether some brands are more expensive on average - whether package size is associated with price - whether categorical variables such as fat content, package type, or location matter - whether a regression question is worth asking A poor graph can also mislead. For this reason, every graph in this course should be interpreted carefully. ## Chapters in Part II | Chapter | Topic | |---|---| | [Chapter 5](chapter-05-descriptive-statistics.qmd) | Descriptive Statistics for Food and Agricultural Data | | [Chapter 6](chapter-06-univariate-graphs.qmd) | Univariate Graphs | | [Chapter 7](chapter-07-bivariate-graphs.qmd) | Bivariate Graphs | | [Chapter 8](chapter-08-multivariate-interactive-graphs.qmd) | Multivariate and Interactive Graphs | | [Chapter 9](chapter-09-graphs-to-research-questions.qmd) | From Graphs to Research Questions | ## Dataset and software All code examples are written for Google Colab. Students should place the dataset in this folder structure: ```text MyDrive/ NREC4107/ data/ Milk_Data_Cleaned.csv notebooks/ outputs/ ``` The code shown on the website is for learning and should be run in Google Colab, not on the website. ## What students should be able to do after Part II After completing Part II, students should be able to: - describe numeric and categorical variables - create basic summary tables - produce univariate, bivariate, and multivariate graphs - identify unusual values and possible data problems - distinguish visual patterns from causal claims - convert visual observations into applied econometric questions ## Key message Exploratory data analysis is not a decorative step. It is part of empirical reasoning. A good graph should help us understand the data and prepare for careful modeling.