yuzaR Data Science on YouTube · Open on YouTube
Foundations: spreadsheets, data literacy and first SQL · Novice · Lesson 9 of 99 · 12 min video, then about 2 hours of practice
Reading data critically: tidy data, grain and data types
After this you can say what one row of a table represents, classify each column's type, and spot a table that will cause wrong answers before you analyse it.
Why it matters
Data literacy is the habit of asking 'what exactly is a row here?' before you sum anything. A table where each row is an order line will double-count revenue if you treat rows as orders. Learn the tidy data rules (each variable a column, each observation a row, each kind of thing its own table) and the difference between nominal, ordinal, interval and ratio data, because that decides which summaries and charts are meaningful.
How to use this resource
Explains the tidy data principles and why messy layouts break analysis. Examples are in R, but the ideas are tool-independent; ignore the code.
Go further
Go deeper