Data and Analysis
Data and Analysis
This domain tests how raw values become usable information: data types and encoding, relational databases and SQL, cleaning and summarizing data, and choosing an honest visualization. Expect items that ask you to interpret a query, spot a misleading graph, or select the right measure of center.
Core rules
- Encoding: text uses ASCII/Unicode code points; images store pixels with color depth; sound is sampled at a rate and bit depth. Lossless compression (PNG, ZIP) preserves exact data; lossy compression (JPEG, MP3) discards detail.
- Relational model: tables have rows (records) and columns (attributes); a primary key uniquely identifies a row; a foreign key references another table's primary key and enforces referential integrity.
- SQL clauses: SELECT chooses columns, FROM names tables, WHERE filters rows, GROUP BY aggregates, HAVING filters groups, ORDER BY sorts.
- Statistics: mean is sensitive to outliers, median resists them; correlation does not establish causation.
Worked example
Given a table students(id, first_name, last_name, grade), return the last names of eleventh graders sorted alphabetically. The query is SELECT last_name FROM students WHERE grade = 11 ORDER BY last_name; WHERE filters before sorting, and only the requested column appears in the result. If the question asked for a count per grade, you would add GROUP BY grade with COUNT(*).
Traps
Swapping primary and foreign keys; using HAVING where WHERE belongs; treating a lossy format as archival; and reading a steep graph as proof of causation when the axes or sample are misleading.