Eli Item-Writing Academy
Most law faculty were never taught to write exam questions — they were handed a template and a deadline. The Academy fixes that: placement by demonstration, sixteen bar-style MBE modules across item craft, MBE formats, application and calibration, and quality at scale, flawed-item drills with triple-checked rewrites, and a completion record you can present for faculty-development documentation.
Unofficial. The Academy is an independent faculty-development resource — not affiliated with the NCBE, the ABA, or any state bar, and not a certification or compliance program. It supports faculty-development documentation; it declares nothing about any school.
Placement by demonstration
The Academy does not ask how good you think your questions are. Tell us what you teach, then take the 60-second drill: six flawed MBE-style items, ten seconds each, one job — name the flaw. Where you land decides your starting module.
Try it now
This is what Academy practice feels like: a flawed item, your diagnosis, then the correction and a triple-checked rewrite. These three are free; the modules hold dozens more.
A seller emails a buyer: "Will sell you my boat for $20,000. Reply by Friday." The buyer replies Thursday: "I'll take it for $18,000." The seller does not respond. Friday passes.
There is no question here, so there is nothing to answer — each option belongs to a different kind of legal conclusion. An unfocused stem turns the item into a guessing game about what the author wanted.
Rewrite (cold-solve verified): add the call — "Is there an enforceable contract?" — and rebuild the options as parallel holdings with reasons. Key: C — the buyer's reply varying the price was a counteroffer that terminated the original offer.
A driver, looking at his phone, fails to notice stopped traffic and rear-ends the plaintiff's car. Which theory of liability applies?
"All of the above" lets an examinee who can eliminate any two options answer without evaluating the third — it tests test-taking, not torts. The fix: replace D with a fourth plausible-but-wrong theory, so every option must be evaluated on content.
Rewrite (cold-solve verified): D. Contract, because the drivers shared the road under traffic laws. Key: C — the driver failed to exercise reasonable care.
A professor sees pooled stats for one of her MBE-style items (112 responses; illustrative data): p-value 0.93, discrimination index −0.16, option pull A 3%, B 2%, C 93% (key), D 2%. What is the correct next step?
A negative discrimination with a very high p-value says the strongest students are missing an easy item — that smells like a miskey or a subtle flaw. But data never diagnose; they only flag. The rule: negative discrimination = faculty review, never auto-delete or auto-rekey.
The curriculum
Work in order within each group — Foundations before Flaw Clinic, the single-best-answer workshop before the linked-sets module — and finish with the Capstone, where you write, peer-review, and submit a ten-item set through Question Studio. Every module ends with a completion record you can download from your account page.
Stem construction, the flaw taxonomy, and distractors that do real work.
Hands-on builds for the MBE's single-best-answer format.
Aiming items at the thinking the MBE measures — and at the right difficulty.
Statistics, fairness, blueprints, peer review — and the capstone set.
Completion records
Finishing a module records it on your device — module, track, level, and date. Your account page shows the full record with a one-click download, ready to present for faculty-development documentation. Records live on your device and travel with the standard account export; one-tap delete removes them like everything else.
What the record is not: it is not a certification, not a compliance statement, and not an endorsement by any accrediting body. It documents that the work was done — nothing more, nothing less.
For students
Students: write your practice questions in Question Studio. Drafting your own MBE-style questions is one of the strongest ways to learn a topic — and every item you write runs through the Studio's cold-solve check, so you never drill yourself on a wrong key. Every item is solved cold twice before it counts; disputed items come back flagged with reasons, never silently shipped.