Instructor discussion questions

Chapter 7 Discussion Questions

  1. Chapter 7 argues that data selection is a design decision, not a neutral technical step. How does this change what students should be asked to justify before they ever choose colors, symbols, or interactions?
  2. Semantic search allows students to find datasets by describing intent, audience, geometry, fields, and narrative purpose. What new forms of responsibility emerge when dataset discovery becomes promptable rather than purely keyword-driven?
  3. Dataset fit can matter more than raw precision. When might a less precise but better-scoped dataset communicate more honestly than a highly detailed dataset that does not match the map's purpose?
  4. A merged dataset can look authoritative because it combines several sources. What evidence should students provide to prove that their join keys, crosswalks, aggregations, and rejected records did not quietly change the map's meaning?
  5. If a student uses an LLM to find, evaluate, clean, and merge data, what should count as authorship, what should count as provenance, and what should count as plagiarism or unacceptable outsourcing?