Categorization¶

Strategy · All strategies
Evidence · 3 claims (3 for) · 7 studies (4 causal, 2 review, 1 quant-synthesis), q2–q4 · 1 of 7 report an effect size

Description¶

Categorization asks learners to actively sort, group, or label cases — problems, examples, specimens, arguments — according to their deep structure rather than surface features. Instead of reading a taxonomy, learners construct it, deciding which items belong together and articulating why.

Design Implications¶

Expertise research shows that what distinguishes experts is not only richer knowledge but better-organized knowledge: experts categorize problems by solution-relevant principles while novices categorize by surface features Chi, Feltovich, & Glaser, 1981 [+S]. Having learners sort cases before instruction primes them to notice the dimensions that matter, preparing them to learn from subsequent explanation Schwartz & Bransford, 1998 [+S]. Sorting tasks also impose manageable load because learners compare items side by side rather than processing each in isolation Chunking reduces working memory load. [+S].

Context¶

Requirements¶

• A case set with contrasting surface features but shared deep structure (and ideally matched surface features across different structures, so surface similarity is not a valid sorting cue)
• A sorting task that requires a decision, not just recognition (Case Studies work well as source material)
• A "time for telling" afterward: instruction that names and explains the categories learners have begun to discriminate Schwartz & Bransford, 1998 [+S]
• Feedback on the correctness of groupings (Assessment or instructor review)

Constraints¶

• Sorting without subsequent instruction leaves learners with intuitions but no principled vocabulary; the contrast activity alone underperforms contrast-plus-lecture Schwartz & Bransford, 1998 [-M]
• If all cases share surface features, learners sort on irrelevant attributes and the activity reinforces misconceptions [~M]
• Poorly differentiated cases (too few, too similar) produce shallow categories; multiple varied representations are needed for flexible category knowledge Cognitive flexibility theory benefits from multiple cases. [+M]

Implementation Variability¶

• Card-sort or drag-and-drop formats (physical or digital) with self-generated category labels
• Problem-indexing: learners label a set of problems with the principle each requires before solving any
• "Odd one out" variants that force fine-grained discrimination
• Progressive sorting: re-sort the same set after instruction to make conceptual change visible

Target Learners¶

• Novices who default to surface-feature reasoning and need to discover structural dimensions Chi et al., 1981 [+S]
• Intermediate learners consolidating a domain's taxonomy before advanced coursework
• Less effective for learners who already possess well-structured category knowledge, for whom sorting is redundant [~W]

Target Learning Goals¶

• Conceptual understanding: discriminating when concepts apply versus when they do not
• Transfer: mapping new problems to known problem types Analogical reasoning improves transfer. [+M]
• Preparation for future learning: building intuitions that make subsequent instruction more productive Schwartz & Bransford, 1998 [+S]

Instructions¶

1. Assemble 8–15 cases that vary surface features while crossing category boundaries (Case Studies, worked problems, or images).
2. Ask learners to group the cases and name each group, working in pairs or small groups to force discussion (Collaborative Learning).
3. Have groups compare their groupings and justify boundaries (Comparing Cases or whole-class discussion).
4. Deliver instruction that names the expert categories and explains the discriminating features — the "time for telling."
5. Ask learners to re-sort the original set, then apply the categories to new cases (Application).

Related Strategies¶

• Comparing cases — the pairwise comparison that categorization generalizes to whole sets
• Concept attainment — the induction-based approach to teaching a single concept via positive and negative instances

Examples¶

• Physics problem sorting — Chi et al.'s classic paradigm: novices sort mechanics problems by "inclined plane" or "pulley"; experts sort by conservation of energy or Newton's second law. Replicating this as a pre-instruction activity reveals and reshapes novice intuitions.
• ISLE Physics curriculum — Students observe contrasting experiments, induce patterns, and categorize observations before formal laws are introduced, following the predict–observe–explain–induction sequence.
• Medical education — Diagnosis teaching using case sets in which similar presentations map to different conditions, forcing learners to categorize by underlying pathology rather than symptom surface.

Key Sources¶

• Chi, M. T. H., Feltovich, P. J., & Glaser, R. (1981). Categorization and representation of physics problems by experts and novices. Cognitive Science, 5(2), 121–152. doi:10.1207/s15516709cog0502_2
• Schwartz, D. L., & Bransford, J. D. (1998). A time for telling. Cognition and Instruction, 16(4), 475–522. doi:10.1207/s1532690xci1604_4
• Marton, F. (2014). Necessary conditions of learning. New York, NY: Routledge.
• National Research Council. (2000). How people learn: Brain, mind, experience, and school (expanded ed.). National Academy Press.