This book is Work in Progress. I appreciate your feedback to make the book better.

Chapter 2 Structured Data

Rows, units, time, and where they live

Some data arrives already rectangular: rows, columns, a value in every cell. The work then is not to build the table but to understand what a row means — one person, one person-year, one month, one country. A surprising share of the mistakes in applied research can be traced back to a misunderstanding on exactly that point.

This chapter walks the three shapes a rectangle can take. A cross-section holds many units at one moment. A panel holds many units at many moments. A time series holds one unit across many moments. They look alike on screen and demand entirely different analysis.

A fourth section asks what the first three take for granted: where does the rectangle actually lie? Frequently not in a file on your own machine, but on a server — too large to copy, or shared by too many people. The habit that follows from this is worth learning early: stop moving data around, start sending questions to it.