ReasonSpan™ · Content quality
Know what has been checked.
The PE Mechanical: HVAC and Refrigeration development preview has completed an AI-assisted subject-matter audit and automated QC. Independent review by a credentialed PE specialist remains pending.
Updated October 18, 2026 UTC. Current assessment edition 2026-10-18.2; transfer edition 2026-10-18.1. Paid enrollment remains closed.
Review scope
- All 24 theory lessons and 24 specialist extensions: equations, worked examples, physical assumptions, units and handbook routes.
- 440 original program questions: diagnostic, two mocks, targeted practice, transfer checks, lesson checks and sampler. Numerical results are recalculated from the printed givens; supplied safety criteria stay within their stated scope.
- Six HVAC Decision Lab workshops, their twelve original decisions, targeted repairs and six delayed follow-up checks. See the shared workshop review status.
- Automated grading, answer withholding, saved-attempt editions, account isolation, retry handling and server-controlled follow-up timing.
AI-assisted review can find errors and improve instruction. It does not substitute for independent professional judgment, candidate trials or measured browser accessibility.
Material corrections from this audit
Edition 2026-10-18.2 adds 96 targeted applications and changes 16 Mock B governing tasks. The separate 28-exercise Chart and Exam Skills Lab adds chart reading and alternative-response training. Historical attempts retain original content and grading. Per-item pacing budgets are author estimates within the working limit; candidate calibration remains pending.
| Location | Finding | Correction |
|---|---|---|
| Assessment whiteboard | Drawing could read an expired React pointer event and blank the entire studio. Reproduced in section two of a saved diagnostic. | Read coordinates before queuing state updates; handle lost pointer capture, secondary pointers and empty strokes safely. Added a regression test for the reproduced failure. |
| Numerical wording and answer cues | Approximate multiple-choice prompts used inconsistent wording; opening answer sequences and four precision differences could cue responses. | Applied “most nearly” to 241 quantitative diagnostic, mock, targeted, sampler, lesson and transfer items. Retained explicit rounding, integer occupancy and catalog-selection requirements. Varied option positions, including multiple-response combinations, equalized four precision cues and replaced six weak distractors with calculated engineering mistakes. That correction was released in edition 2026-10-18.1; new starts now use edition 2026-10-18.2. |
| Saved targeted practice | Checked practice did not record its question edition, making future option reordering unsafe. | Versioned practice responses and retained the original choices, worked review and grading for existing checks and browsers left open during a release. Full attempts retain their original editions. |
| L4 mixing extension | An inlet enthalpy and specific-volume pair could not represent the assumed sea-level vapor-only state. | Replaced the states, specified a common pressure and verified saturation before recalculating the mass-weighted result. |
| D4 flash extension | The supplied liquid and latent enthalpies did not form a realistic saturated-water pair. | Specified receiver pressure and a compatible rounded property pair; recalculated flash and remaining liquid mass. |
| Training clock | Background and closed-browser intervals could inflate study time. | Added foreground presence, a 30-second server expiry safeguard, a 15-minute inactivity pause and a timer-only reset preserving answers, notes, flags and question position. Timed simulation remains continuous. |
| Candidate review of assessment prompts | Several items supplied a standard equation for direct substitution, and some multi-step prompts coached the solution method. | Screened all 240 diagnostic and mock items; revised 51 question presentations, including 16 applied calculation replacements and 32 prompts with reduced method coaching. Reviewed rounding, dimensionless labels and dew-point distractors. That correction was released in edition 2026-10-15.1; original saved questions remain intact. |
| Saved-attempt equipment figures | Older visual editions still displayed basic equipment boxes. | Added cooling-tower, motor, pump and flash-receiver 3D illustrations with exact vector schematics. Extended compatible archived-edition presentation mappings, including sensible recovery. |
| Diagnostic and mock numeric choices | A diagnostic LMTD choice equaled a rounded correct result; two mock LMTD distractors and several other alternatives were too closely spaced. | Revised nine items with distinct, recalculated engineering errors; made diagnostic LMTD precision explicit. That correction was released in assessment edition 2026-10-14.1. Existing attempts retain their original choices. |
| First-section diagnostic figures | Five generic box diagrams did not adequately connect equipment to exam schematics. | Added realistic equipment companions and precise vector schematics with legible labels, unchanged givens and no computed answers. Compatible saved attempts receive the presentation update. |
| Diagnostic dew-point item | The supplied saturation pressures did not match their labeled temperatures. | Corrected the pressure/temperature data. Started assessments retain their original edition. |
| L4 transfer check | The outdoor-air properties implied an implausibly hot design state at sea level. | Used realistic inlet states at an explicit common pressure; updated the calculation and schematic. Saved checks retain the old question and figure. |
| D3 transfer check | The suction pressure location was ambiguous, and excess pump head needed an operating assumption. | Named the upstream pressure node, elevation datum and intervening loss. Specified balancing for excess head at the required flow. |
| Cold-air interpolation workshop | Synthetic lookup values could be mistaken for approximate real ice properties. | Explicitly labeled them hypothetical interpolation data. Real frost prediction requires accurate saturation-over-ice properties. |
| Charge and airflow distractors | Two alternatives lacked a clear, useful engineering error. | Replaced arbitrary values with an incorrect mass comparison and an incorrect standard-density substitution. |
| Controls follow-up | The earlier follow-up repeated a protective-state conclusion too closely. | New starts test verified release and the complete normal command set. Previously started follow-ups retain their question. |
| D6 handbook route | The route emphasized a general system heading and the valve coefficient alone. | Added the handbook's control terminology, control types, valves and authority sections. |
Revised mock practice
The current edition replaces 48 questions in each mock with multi-step problems, two per module. The audit checks printed givens, independent numerical results, feasible physical states, meaningful distractors, duplicate stems, domain coverage and legible figures and tables. Shorter questions remain for pacing.
Saved attempts preserve the original question edition. Automated checks and AI-assisted technical review do not establish exam-equivalent difficulty; independent specialist solving and candidate calibration remain pending. Read how the mocks rehearse the exam process.
Module review status
What remains to be validated
Independent blind solving and distractor review by a qualified specialist; observed candidate trials; and broader production browser, mobile and accessibility acceptance. A focused production check reproduced the section-two whiteboard failure; the correction is covered by a deferred-event regression test. Record the reviewer, qualifications, date and resulting corrections before claiming independent review.
The NCEES handbook is the exam reference; the current specification lists no separate design-standard PDFs. Expected knowledge still extends beyond the handbook. Original practice criteria are not universal code rules or approval for a real installation.
Scores and confidence help choose the next study task. They are not calibrated mastery measures or pass predictions. Reflections and repair completion are self-reported.