Assessment Realism Without Counterfeit Authority

SECTION 01

Pin the official specification as executable constants

The release represents 130 scored items, 10 pretest items, 180 minutes, and an integer passing score of 92, with exact domain quotas.

Assessment realism starts with a versioned specification, not visual imitation. The current release defines scored quotas of 20 economic-and-business items, 32 investment-vehicle items, 39 client-recommendation-and-strategy items, and 39 laws-regulations-and-ethics items. These sum to 130. A simulation adds ten eligible unscored items and stores a 180-minute expiration. Results compare earned scored answers directly with 92.

Every number is both displayed and tested. The practice page explains the blueprint and links to NASAA sources; the selector enforces quotas; the database stores the threshold on the session; results label unscored teaching items separately. Pinning values in one assessment module limits drift, but maintainers still must verify the official outline and study guide before a future release. A constant is evidence of implementation, not evidence that an external rule can never change.

SECTION 02

Approximate the blueprint honestly for short practice sets

Short sets use largest-remainder allocation, then concept-diverse selection, while focused filters explicitly opt out of blueprint proportions.

A five-question set cannot reproduce fractions such as 20/130 exactly. The allocator multiplies the requested count by each quota, takes every floor, and distributes leftover questions by descending fractional remainder with stable domain-order tie breaking. That yields exact whole-question totals and makes the approximation inspectable. The UI prints the resulting mix before launch rather than describing every mixed set vaguely as blueprint aligned.

Selection then groups eligible questions by concept, favors the least-represented groups, randomizes within groups, and performs a final shuffle that avoids adjacent repetitions when possible. If a learner selects a domain, lesson, difficulty, missed-only pool, or bookmarks, the interface switches to focused remediation language. That distinction matters: usefulness for a targeted weakness and resemblance to a full blueprint are different optimization goals.

exact = requested × domainQuota ÷ 130
base  = floor(exact)
leftovers → largest fractional remainders
final counts → exact requested total

SECTION 03

Exercise simulation mechanics without advertising a launch

The backend and E2E suite support full simulation sessions, but the learner practice page intentionally exposes only targeted practice.

A simulation session can load either of two release-pinned 140-item templates, randomize display order within each item's four options, defer all feedback, allow answer changes and review marking, enforce expiration, and grade only at finish. Browser tests create such a session through the same-origin session endpoint, then drive the real question runner and results pages. This verifies the mechanics without creating a competing pseudo-official product promise.

The capability boundary is explicit: simulations are exercised through backend and E2E paths, but there is no learner-facing simulation launch control. The practice page contains no Start simulation button and no Series 65 exam-conditions card. A test asserts both absences. This is intentional product scope, not an unfinished button: targeted practice is the public learner entry point while full-form behavior remains testable for correctness and future review.

SECTION 04

Make originality and non-endorsement impossible to miss

The portal mirrors mechanics and topic weights, not confidential questions, official score reports, or regulator approval.

All portal questions are original study items connected to public source metadata. NASAA does not release live questions or endorse exam-preparation providers. The portal therefore avoids claims such as official question, actual exam item, guaranteed pass, or NASAA approved. It can accurately say that this release applies the published blueprint, timing, scored total, and passing boundary to original educational material.

The results page preserves the distinction after the most realistic interaction. Its item-by-item review and visible unscored labels are teaching features; the page states that an official NASAA score report does not disclose question text or identify pretest items. Domain cards are diagnostics and are not averaged to determine the total. A 92/130 mock result means this original form met the pinned threshold under portal mechanics, not that NASAA administered, reviewed, or endorsed the attempt.

• Use “original, blueprint-aligned study item,” never “official question.”
• Describe 80% on short practice as a study target, not an official score.
• Describe 92/130 as the pinned official scored threshold represented by the mock mechanics.
• State that NASAA does not endorse exam-preparation providers and does not release live questions.