Case Comparison and Contrast¶

Comparison method — instantiates Transfer Scaffolding

Sets several cases side by side so the learner induces the invariant they share and reads off the boundaries where it stops holding, instead of over-fitting to any single example.

A single example teaches its surface as much as its structure; several examples, compared, teach the structure alone. Case Comparison and Contrast sets multiple cases side by side and has the learner induce the invariant they share and read off the systematic differences that mark where it breaks. Its defining move — distinct from a one-to-one analogy — is that the comparison itself does the abstracting: laying cases together forces out both the transferable principle and its boundary conditions, which no single case can reveal on its own.

Example¶

An analyst assessing "run" risk in stablecoins compares four historical episodes side by side: the 1907 Knickerbocker panic, a 1980s savings-and-loan failure, a money-market fund "breaking the buck," and a 2020s algorithmic-stablecoin collapse. Aligned, the invariant surfaces: a promise of redemption at par + a maturity or liquidity mismatch + a coordination trigger → a self-fulfilling run. But the contrasts are just as informative: episodes with a credible lender of last resort, hard redemption gates, or genuinely full reserves behaved differently — those are the boundaries where the pattern weakens or switches off.

The output is a portable principle (run risk ≈ redemption promise × mismatch × absence of a circuit-breaker) together with the conditions that neutralize it. The analyst can now size up a new stablecoin never seen before, because the schema came from many cases rather than being over-fit to one.

How it works¶

The method is juxtaposition plus a forced inference:

Place two or more cases side by side, aligned on comparable elements.
Ask what is the same across them — the candidate invariant — and what differs systematically.
Promote the shared relational structure to a stated principle, not a shared surface feature.
Promote the systematic differences to boundary conditions: the circumstances under which the principle stops holding.

Tuning parameters¶

Number and spread of cases — more varied cases sharpen the invariant and expose more boundaries, at more effort. Too-similar cases yield a narrow or false principle.
Alignment scaffolding — how much the cases are pre-aligned for the learner versus aligned by the learner (harder, but deeper).
Same- vs contrasting-outcome cases — cases that diverged in outcome are the cheapest way to find boundary conditions.
Abstraction push — stop at listing similarities, or force an explicit stated principle.
Case order — matched pairs versus interleaved, which changes what contrasts become salient.

When it helps, and when it misleads¶

Its strength is that comparison produces genuinely portable schemas where single examples produce surface-bound memories, and contrasting outcomes are the most economical route to boundary conditions.^[1]

Its failure mode is bad case selection: cherry-picked or near-identical cases yield a false or too-narrow invariant, and learners readily latch onto a salient surface commonality instead of the relational one. The classic misuse is assembling only confirming cases — none that break the preferred pattern. The discipline is to include at least one case that did not follow the pattern and to keep the relational invariant separate from any shared surface feature.

How it implements the components¶

Case Comparison and Contrast fills the induce-and-bound subset of the archetype:

shared_structure_map — the invariant induced across the cases: the transferable principle, expressed relationally.
boundary_condition_set — the systematic differences the contrasts expose, marking where the invariant stops holding.

Where Analogical Mapping Prompt maps one source to one target, this induces from many; it does not turn its boundaries into an enforced readiness gate (Boundary Condition Checklist's transfer_readiness_check consumes them) or provide any target practice.

Instantiates: Transfer Scaffolding — it derives the transferable principle and its limits from a set of examples.
Sibling mechanisms: Analogical Mapping Prompt · Boundary Condition Checklist · Bridging Examples · Worked Example Variation · Negative Transfer Red Team · Apprenticeship Reflection · Cross-Context Practice · Implementation Coaching · Near-to-Far Transfer Sequence · Transfer Debrief · Transfer Prompt

References¶

[1] Analogical encoding (Gentner and colleagues) — comparing two examples against each other promotes abstraction of their shared schema far more than studying the same examples one at a time. Contrasting cases with different outcomes is the same move aimed at the schema's limits. ↩