Ambiguous tasks with a revision step help students recognize the role of assumptions in mathematical activity¶

Claim · All claims
Evidence · 1 study · design r3 · q3 · n=2 fully analyzed episodes (38 grade-9 students; 22 grade-5 students), from 5 classes across 3 design cycles over 3 years, Japan
Currency · newest study 2024 · not yet checked against its sources · not reviewed by a person

Both grade 5 and grade 9 students, given a task with a deliberately unspecified local assumption, produced genuinely divergent, individually defensible answers, and through structured discussion and a task-revision sub-task, articulated both that a conclusion's truth depends on unstated assumptions and that reaching a shared answer requires making those assumptions explicit.

Subclaims¶

q3 i? In a grade 9 class of 38 students given two data points and asked to find a missing value, all students in the class-wide discussion spontaneously agreed that two different answers (y=6 and y=-6) were both correct because each depended on a different, unstated assumption about the underlying functional relationship. → Komatsu et al. 2024

q3 i? In a grade 5 class of 22 students asked to divide a rectangle into "four same parts," initial individual responses split 17-to-4 in favor of a single interpretation (congruent triangles), but discussion led several additional students to explicitly recognize a second interpretation (equal area) as legitimate, with one student's final written reflection stating "different people considered part differently... I thought it was absolutely Option 1, but I listened to others' opinions and realized that there are other ways of thinking." → Komatsu et al. 2024

q3 i? In both episodes, a follow-up sub-task asking students to revise the task's wording so only one answer remained correct led students to propose and evaluate increasingly precise formulations (e.g., moving from "divide into triangles" to "divide into four congruent triangles"), and to state explicitly, in their own words, that specifying an assumption is what makes a unique answer possible. → Komatsu et al. 2024

q3 i? One teacher's informal follow-up observation (not part of the systematic data collection) suggested the effect might persist beyond the single lesson: students from the study's class, given an unrelated ambiguous problem later in the year, spontaneously identified the missing assumption, whereas a comparison class not in the study did not. → Komatsu et al. 2024

Evidence¶

Komatsu et al. 2024¶

Komatsu, K., Murata, S., Stylianides, A. J., & Stylianides, G. J. (2024). Introducing Students to the Role of Assumptions in Mathematical Activity. Cognition and Instruction, 42(2), 327-357. https://doi.org/10.1080/07370008.2023.2293695

q3 · peer-reviewed design-based research across three iterative cycles and five classes (two secondary, three elementary), with systematic bottom-up and top-down coding of two fully-analyzed episodes against a hypothetical learning trajectory; not a controlled experiment i? · the abstract prints no effect size; the full text may n=2 fully analyzed episodes (38 grade-9 students; 22 grade-5 students), from 5 classes across 3 design cycles over 3 years, Japan · design · r3

Students in two analyzed classrooms — one secondary (functions), one elementary (geometry) — worked on tasks designed via Productive Ambiguity Task Design for Assumptions. Video, transcripts, and student worksheets were coded first bottom-up (open coding of lesson episodes) and then top-down (comparison against a hypothetical learning trajectory). Both classes reached the two targeted learning goals — recognizing that a conclusion depends on assumptions, and recognizing that explicit assumptions are needed for a shared answer — through direct engagement with the ambiguity and the task-revision sub-task, at both grade 5 and grade 9.

Discussion¶

This is design-based research across a small number of intensively-analyzed classes (2 detailed episodes from 5 total classes), not a large-sample or controlled study; effects are documented through direct quotation and coding of classroom discourse and written reflections, not standardized assessment. The one report of a possible lasting effect beyond the single lesson is explicitly anecdotal (a single teacher's informal comparison, not part of the systematic data collection) and should be weighted accordingly. The authors also note a boundary condition: too much prior knowledge of related mathematical variability (e.g., familiarity with many function types) could cause students to conclude a task is simply unanswerable rather than productively comparing specific interpretations, so task design must be calibrated to students' existing knowledge.

Related Claims¶

• Students' argumentative essay quality improves substantially from initial to revised drafts regardless of feedback source — the general claim this finding bears on, with the studies behind it
• (none yet linked)
• Minimal guidance is less effective for novices than explicit instruction — related
• Before the intervention, most students could not write assumptions and conclusions, and only one student achieved the maximum pre-test score — related
• Underlying assumptions dictate how the five foundations are operationalized in a learning environment — related
• The authors argue that analyzing power imbalances, misshapen structures, and hidden assumptions are familiar critical thinking tasks for hesitant readers — related
• Material difficulty is not determined solely by element number and interactivity — a broader claim this one bears on