Real tables arrive with missing values, duplicate rows and inconsistent formatting, and cleaning them is usually the first half of any analysis. NULL is not zero and not an empty string, so it needs IS NULL and COALESCE rather than an equality test, and duplicates are removed most reliably by ranking rows with ROW_NUMBER and keeping the first of each key. These questions give you dirty seeded data and ask for the clean result set.
What you will practice
•Test for missing values with IS NULL and supply defaults with COALESCE or NULLIF
•Deduplicate with ROW_NUMBER() OVER (PARTITION BY key ORDER BY updated_at) and keep row number 1
•Understand why NULL breaks equality comparisons, NOT IN and some aggregate counts
•Normalise inconsistent text with TRIM, LOWER and CASE mappings before grouping on it
•Validate a cleanup by comparing row counts and distinct key counts before and after
