Linked Sets: Multi-Question Fact Patterns

Learning goals

Decide when a linked multi-question set serves learning better than standalone items.
Write sets where each question stands on its own facts-with-context (no answer-chaining).
Keep every question in a set mapped and cold-solvable independently.

Lesson 1 — What Linked Sets Are (And Aren'T)

The MBE itself is all standalone items — but in practice banks, linked sets (several questions sharing one fact pattern) are a legitimate teaching format: they simulate working a matter from multiple angles. What they are not: a way to make one question's answer depend on another's. Each question in the set must be answerable from the shared pattern alone.

Lesson 2 — The Independence Rule

The cardinal sin of linked sets is answer-chaining: Q2's key depends on getting Q1 right. If Q1 asks "is there a contract?" and Q2 asks "what are the damages for the breach?", an examinee who misses Q1 is locked out of Q2 — and you've measured one thing twice. Write each question so a solver who skipped the others could still answer it.

Lesson 3 — Vary The Tested Point

A linked set that asks the same doctrine four ways is one item with extra steps. Use the shared pattern's richness: the formation question, the defenses question, the remedies question, the parol-evidence question — each mapped to its own outline subtopic, each with its own key and distractors.

Lesson 4 — Keep Sets Short

Two to four questions per pattern. Longer sets exhaust the pattern's facts (later questions start testing trivia) and punish examinees who misread one early sentence. When the pattern can't support another independent question, end the set.

Spot-the-flaw drill

Flaw: Q2 chains on Q1 — its premise ("if the contract from Q1 is enforceable") require

Q2 chains on Q1 — its premise ("if the contract from Q1 is enforceable") requires the Q1 answer, so Q2 measures nothing independently and confuses the solver who got Q1 right (there is no contract, so the hypothetical is incoherent).

Correction

Rewrite Q2 to stand alone on the shared facts: ask about a different doctrine the pattern supports (e.g., whether the Thursday reply could operate as an acceptance under any theory), or extend the pattern with new independent facts.

Model rewrite

— cold-solve verified: 2/2 on each question independently

Shared pattern: (same as above).

Q1: Is there an enforceable contract for the boat? Key: No — the Thursday reply was a counteroffer that terminated the original offer, which then lapsed on Friday.
Q2 (independent): Suppose instead the buyer had replied Thursday: "I accept. Will pick up Saturday." The seller then sells to someone else Friday morning. Is the seller in breach? Key: Yes — a contract formed on acceptance Thursday; the Friday sale is a breach (anticipatory/repudiation analysis).

Drill key

Q2 now varies the facts into an independent tested point (acceptance + breach) rather than chaining on Q1's answer. Both solvers answered each question independently and converged on both keys.

Sources

Downing, S. M., & Haladyna, T. M. (Eds.). (2006). Handbook of test development. Lawrence Erlbaum Associates. (On testlet design and local item dependence.)