Goal: preserve known second-variable data when the first label is missing. 25 = 2 + 5 + 7 + 7 + 4.
- 0–2: Source C's eight no-theme designs have known side counts. Missing theme is not missing side count.
- 2–7: Remove that row as a demonstration only: named-theme two-sided count is 27 of 42 (≈64.3%); restore the row for the all-design claim of 30/50=60%. Both fractions answer different questions.
- 7–14: Learners complete a two-column claim audit: 27/42 named-theme designs versus 30/50 all designs. Circle the excluded three two-sided and five one-sided designs.
- 14–21: Routes: physically remove/restore eight marked cards; shade the no-theme table row and write both fractions; dictate a two-sentence comparison with denominator labels. No route guesses why labels are absent.
- 21–25: Exit “Could 27/42 be described as ‘all designs’?” Look-for: no; eight were excluded. Next move: if learner treats no-theme as zero, restore its 3 and 5 cells.
Another setting: Imaginary theatre costume sketches with undecided colour but known sleeve type. Optional/home: explain to a fictional designer why unknown in one field need not erase a row.