Dialogic teacher support cultivates all four aspects of statistical modeling practice over a multi-week unit¶

Claim · All claims
Evidence · 1 study · qualitative r2 · q2 · i1 small · n=28 sixth-grade students (modal age 11, 72% eligible for lunch supplements), one classroom, 4 weeks (15 lessons)
Currency · newest study 2024 · not yet checked against its sources · not reviewed by a person

Across a 4-week statistical modeling unit taught by a teacher with no specialized statistics training, nearly all identified classroom episodes involving one of four foundational modeling aspects also included an explicit, identifiable form of dialogic teacher support for that aspect, and students showed qualitative progress on all four aspects by the unit's end.

Subclaims¶

q2 i1 Of 182 classroom episodes coded across 15 lessons, 179 (98%) included at least one explicit form of teacher assistance aligned with one of the four modeling aspects (81 episodes involved mapping situation to model, 58 simulating a long-run process, 34 evaluating model fit, 47 making model-based inferences; 24 episodes involved more than one aspect). → Wisittanawat & Lehrer 2024

q2 i1 Students' language and reasoning shifted to reflect each aspect over the unit: for mapping, most students moved from assuming equiprobable outcomes in a two-outcome chance situation to representing an actor's actual historical probability (e.g., correctly modeling a 70% free-throw shooter) after repeated teacher prompts highlighting the actor's history; for inference, students moved from purely visual judgments ("that looks like an outlier") to quantified probability language ("5% chance," "20 times more unusual"). → Wisittanawat & Lehrer 2024

q2 i1 On an external post-unit assessment, most student responses about samples reflected a hierarchical conception — treating a sample as both a specific set of outcomes and one member of a potentially infinite collection of such samples — consistent with the long-run-process aspect of statistical modeling practice. → Wisittanawat & Lehrer 2024

Evidence¶

Wisittanawat & Lehrer 2024¶

Wisittanawat, P., & Lehrer, R. (2024). Teacher Cultivation of Classroom Statistical Modeling Practice: A Case Study. Cognition and Instruction, 42(4), 505-547. https://doi.org/10.1080/07370008.2024.2397969

q2 · intensive single-classroom qualitative case study with systematic thematic coding (constant comparison) of 182 classroom episodes across 15 fully-observed lessons; a participant-observer researcher was present for all lessons; no comparison classroom or quantified pre/post learning measure beyond an external post-unit assessment described qualitatively i1-i2 · a clear, well-documented mechanism (near-universal presence of aligned teacher support) and consistent qualitative evidence of student progress on all four aspects, achieved by a non-specialist teacher, though without a quantified effect size or comparison condition n=28 sixth-grade students (modal age 11, 72% eligible for lunch supplements), one classroom, 4 weeks (15 lessons) · qualitative · r2

A sixth-grade teacher with no specialized statistical training taught a 4-week unit using Dialogic Facilitation of Statistical Modeling Practice, covering all 28 students in the class. A researcher present as participant-observer for every lesson video-recorded and field-noted all whole-class and small-group talk; 182 episodes were coded via constant comparison against the four modeling aspects, with recurring forms of teacher assistance identified for each. Student reasoning was tracked qualitatively across the unit and via an external post-unit assessment of sampling understanding.

Discussion¶

This is a single-classroom case study with one teacher and no comparison condition, so the claim documents feasibility and mechanism (what dialogic support for statistical modeling looks like, and that a non-specialist teacher can enact it) rather than a measured effect size or generalizable prevalence across teachers. The coding of "explicit teacher assistance" in 98% of episodes is a descriptive frequency from the same dataset used to develop the coding scheme, which likely inflates apparent alignment somewhat; the qualitative evidence of student progress (e.g., the 70%-free-throw example) is illustrative, drawn from specific episodes rather than assessed across all 28 students systematically.

Related Claims¶

• (none yet linked)
• Positioning students as sources increases productive participation in science discourse — related
• Subtle teacher guidance, not imposition, enables interdisciplinary integration in small-group deliberation — related
• Heterogeneity-seeking modeling curricula surface epistemic commitments in most students, including care-based commitments absent from prior frameworks — related
• Teachers play a key role in engaging students in writing and improving writing outcomes in grades 6-12 — related
• Qualitative coding of detected clips found students most often gazed at task materials (47.6%), overt help-seeking was nearly absent (1.6%), and the teacher was present in 49.2% of segments — related