Math Talks¶

Strategy · All strategies
Evidence · 3 claims (3 for) · 8 studies (6 quant-synthesis, 1 review, 1 theoretical), q2–q4 · 6 of 8 report an effect size

Description¶

Math talks are structured conversations focused on mathematical reasoning and problem-solving strategies. A facilitator poses a problem — often one with multiple solution paths — and learners explain their thinking aloud, compare strategies, and work toward shared precision in mathematical language. The instructor's role shifts from delivering procedures to eliciting, probing, and connecting student ideas, which also yields real-time evidence of comprehension.

Design Implications¶

Math talks operationalize Active Learning in mathematics: verbalizing reasoning forces learners to organize and justify their thinking, and exposure to peers' strategies builds flexible problem-solving rather than single-procedure fluency Active learning improves exam performance relative to lecture. [+S]. Their effectiveness depends on discourse norms — learners must be accountable to the mathematical community, not just to the teacher Kazemi & Stipek (2001) [+M]. Facilitator moves matter: pressing for reasoning ("How do you know?"), linking contributions ("How is yours like hers?"), and requiring that everyone understand each viewpoint develop both conceptual depth and mathematical precision.

Context¶

Requirements¶

• A well-chosen problem or scenario with multiple accessible solution paths
• A facilitator skilled in probing questions and in connecting responses without evaluating them prematurely
• A classroom culture that supports open discussion, risk-taking, and respectful critique of ideas
• Norms of accountability: the expectation that explanations make sense to everyone, using precise mathematical terms
• A well-chosen problem with multiple accessible solution paths (mental math strings, dot images, open-ended word problems)
• A facilitator skilled in neutral, probing moves: "How did you see it?", "Who solved it a different way?", "Do you agree or disagree, and why?"
• Recording of learner strategies publicly (board, document camera) so representations become shared objects of discussion
• Established norms of psychological safety — errors are treated as sites of learning, not failures

Constraints¶

• Without careful planning, discussion can drift toward a sequence of show-and-tell answers with no mathematical connection between them Michaels, O'Connor & Resnick (2008) [~M]
• Requires established norms; in classrooms where students expect answer-checking rather than reasoning-sharing, early talks may produce low-quality participation [-M]
• Time-intensive relative to direct explanation; overuse on routine problems yields little conceptual return [~W]
• Dominant voices can suppress participation from less confident learners unless structures (turn-and-talk, written rehearsal, random calling) distribute airtime [-M]
• Fails when the facilitator evaluates answers immediately, collapsing the discussion into Direct Instruction with extra steps [-M]
• Requires sustained culture-building; in classrooms without discussion norms, low-status learners may be excluded and talk remains dominated by a few voices [-M]
• Ineffective when problems have a single obvious path or when time pressure forces the facilitator to truncate the comparison of strategies [-W]
• Whole-class format can leave quieter learners disengaged; paired or small-group variants mitigate this [~M]

Implementation Variability¶

• Number talks: brief, mental-math focused, emphasizing strategy comparison on a single computation
• Whole-class sharing: learners present solutions at the board while the facilitator orchestrates connections
• Peer discussion first: pairs or small groups rehearse reasoning before whole-class sharing (Peer Discussion)
• Written-to-oral progression: learners annotate their work before speaking, lowering the entry barrier for multilingual learners
• Number talks (Parrish, 2010): short mental-computation routines with dot cards or computation strings, common in K–5
• Small-group discourse: learners discuss in pairs before whole-class sharing, increasing participation density
• Student-led variants: learners facilitate the discussion once norms are established
• Online adaptations: chat-based strategy sharing or digital whiteboards (e.g., Desmos Classroom) where learners submit strategies anonymously before discussion

Target Learners¶

• K–12 mathematics learners across grade levels; problem complexity is adjusted rather than the format
• Learners who benefit from hearing multiple solution strategies, including those who solve correctly but cannot yet explain why
• Multilingual learners, when paired with sentence frames and written rehearsal, because math talks make academic mathematical language explicit [+W]
• Less effective for learners who already have flexible, well-connected strategy repertoires, for whom strategy comparison may be redundant [~W]
• K–12 learners at all levels; the routine scales by adjusting problem complexity rather than format
• Learners with weak procedural fluency benefit from seeing multiple efficient strategies modeled by peers [~M]
• Multilingual learners benefit when representations and gestures supplement verbal explanation, though oral-only formats can disadvantage them [~W]

Target Learning Goals¶

• Conceptual understanding: why procedures work, not just how
• Procedural flexibility: comparing and evaluating multiple solution strategies
• Mathematical communication: precision, justification, and critique of others' reasoning
• Formative assessment: surfacing misconceptions for the instructor in real time Feedback is most effective at task and process levels. [+S]
• Conceptual understanding of number, operations, and relationships
• Procedural flexibility: knowing when to apply which strategy, not just how
• Mathematical communication: justifying, critiquing, and using precise language
• Formative insight for the instructor into learner thinking Formative assessment improves achievement. [+S]

Instructions¶

1. Select a problem with multiple solution paths and anticipate the strategies learners will use.
2. Give individual think time before any talk; optionally allow written rehearsal or pair rehearsal (Peer Discussion).
3. Collect answers or strategies publicly first, then have learners explain their reasoning (Eliciting Student Thinking).
4. Press for reasoning with probing questions; record and label strategies so the class can compare them.
5. Orchestrate connections between strategies — ask learners to compare, contrast, and evaluate approaches rather than only presenting their own.
6. Close by naming the key mathematical idea and the precise language that describes it; follow with Practice on related problems.

Related Strategies¶

• Cognitively Guided Instruction — math talks are the discourse engine of CGI: teachers elicit and build on children's informal strategies
• Discussion-Based Learning — the general discussion pattern math talks specialize for mathematics
• Case-Based Learning — shares the structure of discussing and comparing solution approaches to a shared problem
• Formative Assessment — math talks double as a real-time window into learner reasoning
• Extend rich student discourse beyond Math Talks and use participation structures and story problems to broaden access and sense-making

Examples¶

• Number Talks (Parrish, 2010) — A K–5 routine: a computation such as 16 × 25 is posed mentally, learners share strategies ("I did 25 × 4 four times…"), and the facilitator records each on the board for comparison. (Math Solutions)
• Classroom Discussions (Chapin, O'Connor & Anderson, 2009) — A published professional program training teachers in five talk moves (revoicing, repeating, reasoning, adding on, waiting) used across K–12 math lessons.
• Complex word problem debrief — After learners attempt a multi-step word problem, the instructor facilitates a talk comparing a bar-model solution with an algebraic one, pressing each presenter until peers can restate the reasoning in their own words.
• Number Talks routine (Parrish, 2010) — a K–5 teacher poses "18 × 5" mentally; learners share strategies such as (10 × 5) + (8 × 5) and doubling 45, and the teacher records each on the board for comparison. See https://www.mathperspectives.com
• Desmos Classroom (https://teacher.desmos.com) — digital activities that collect every learner's strategy and display anonymized samples for whole-class comparison, structuring the same discourse pattern online.
• Youcubed (Stanford) (https://www.youcubed.org) — publishes low-floor, high-ceiling tasks and teacher guidance designed for discussion-rich mathematics classrooms.

Key Sources¶

• Kazemi, E., & Stipek, D. (2001). Promoting conceptual thinking in four upper-elementary mathematics classrooms. The Elementary School Journal, 102(1), 59–80. doi:10.1086/499693
• Michaels, S., O'Connor, C., & Resnick, L. B. (2008). Deliberative discourse idealized and realized: Accountable talk in the classroom and in civic life. Studies in Philosophy and Education, 27(4), 283–297. doi:10.1007/s11217-007-9071-1
• Webb, N. M., Franke, M. L., Ing, M., Chan, A., De, T., Freund, D., & Battey, D. (2014). Engaging with others' mathematical ideas: Interrelationships among student participation, teachers' instructional practices, and learning. International Journal of Educational Research, 63, 79–93. doi:10.1016/j.ijer.2013.02.001
• Chapin, S. H., O'Connor, C., & Anderson, N. C. (2009). Classroom discussions: Using math talk to help students learn, Grades K–6 (2nd ed.). Math Solutions.
• National Council of Teachers of Mathematics. (2014). Principles to actions: Ensuring mathematical success for all. NCTM.
• Hufferd-Ackles, K., Fuson, K. C., & Sherin, M. G. (2004). Describing levels and components of a math-talk learning community. Journal for Research in Mathematics Education, 35(2), 81–116.
• Parrish, S. D. (2010). Number talks: Helping children build mental math and computation strategies, grades K–5. Math Solutions.
• Freeman, S., et al. (2014). Active learning increases student performance in science, engineering, and mathematics. PNAS, 111(23), 8410–8415. doi:10.1073/pnas.1319030111