10.5 Developing and Administering MT Qualification Examinations

10.5 Developing and Administering MT Qualification Examinations

Develop MT personnel certification examinations and proctor MT exams (written and practical) are separate tasks in the official Level III MT topic list, and they are the tasks where a weak programme does the most damage. A certification is only as strong as the examination behind it; an examination that any attentive trainee can pass without understanding MT converts the whole qualification system into paperwork.

---

The Three-Part Examination Structure

Grading (SNT-TC-1A and CP-189 baseline): a minimum of 70% on each examination and a composite of at least 80%, the composite being the simple average of the general, specific, and practical grades.

$\text{Composite} = \frac{G + S + P}{3} \ge 80\% \quad \text{with} \quad G,\ S,\ P \ge 70\% \text{ individually}$

The classic trap: 95, 95, 68 averages to 86.0% and still fails, because the practical fell below the 70% individual floor. A composite above 80% never rescues a sub-70% part.

For Level III candidates under CP-189, the employer's examination package sits on top of the central ASNT NDT Level III certificate: the employer administers a specific examination on its own codes, specifications, and equipment and a practical examination that includes preparing an NDT procedure.

---

Building the Question Bank

1. Blueprint before writing. Allocate questions across the CP-105 topical outline items for the method and level, in proportion to their weight in the work. A bank that is 60% magnetization formulas and 5% interpretation does not measure an MT inspector.
2. One testable point per item. An item should fail for exactly one identifiable reason. Compound items ("which of the following are true about AC and HWDC and demagnetization?") measure reading stamina.
3. Plausible distractors from real errors. The best distractors are the mistakes technicians actually make: applying the 90 to 110 A/in. prod bracket to a 1 in. plate, using the off-centre coil formula for a centred part, reading total sediment instead of particle volume, quoting 1 minute of dark adaptation under Article 7.
4. Avoid structural cues. Eliminate "all of the above," grammatically mismatched options, longest-option-is-correct patterns, absolutes such as "always" and "never" in distractors only, and answer-position clustering. Spread correct answers roughly evenly across positions.
5. Write explanations that reference content, not letters. An explanation reading "B is correct" teaches nothing and breaks when the bank is reshuffled.
6. Review and retire. Have a second qualified person technically review every item against the current code revision. Track item difficulty and discrimination; retire items that everyone passes, that everyone fails, or that good candidates miss more often than weak ones. Re-verify the bank whenever a cited code revision changes — a correct 2019 item can become a wrong 2026 item, as the yoke verification interval demonstrates.

---

Designing the Practical Examination

The practical is the part most often reduced to a formality, and the part that best predicts field competence. A defensible MT practical specifies:

• Specimens: representative material, product form, and surface condition, containing documented discontinuities, plus at least one specimen carrying non-relevant indications and ideally one sound specimen. Specimen flaw maps are controlled documents held by the Level III, not posted in the lab.
• Tasks: system set-up and performance verification (bath concentration, lighting readings, field adequacy by shim or field indicator), correct technique and current selection for the geometry, execution of both required field directions, interpretation and evaluation against stated acceptance criteria, demagnetization with residual field measurement, post-cleaning, and completion of the actual report form.
• Checkpoint scoring: a written rubric with weighted checkpoints and a stated pass threshold, so two proctors score the same performance the same way. Detection performance and false calls are both scored.
• Time limits and conditions: realistic, stated in advance, identical across candidates.
• Retention: the graded rubric, the candidate's completed report, and the specimen identities are retained in the personnel file.

---

Proctoring, Security, and Re-Examination

+-----------------------------------------------------------------------------+
|                  EXAMINATION INTEGRITY CONTROL CHECKLIST                    |
+-----------------------------------------------------------------------------+
| BEFORE  | Verify candidate identity, current vision record, and that        |
|         | training/experience prerequisites are met and documented         |
|         | Confirm exam version selection; issue only permitted references  |
|---------+-----------------------------------------------------------------|
| DURING  | Continuous invigilation; no unauthorized notes, phones, or       |
|         | reference material beyond what the written practice allows;      |
|         | no coaching by the proctor; log start/stop and any incident      |
|---------+-----------------------------------------------------------------|
| AFTER   | Collect all papers and specimens; grade against the key/rubric;  |
|         | second-person verification of grading arithmetic; record grades  |
|         | in the personnel file; return exam materials to controlled       |
|         | storage; never return a graded master to the candidate           |
+-----------------------------------------------------------------------------+
| BANK    | Multiple equivalent versions; items not reused verbatim as        |
| SECURITY| training handouts; access restricted to the Level III and         |
|         | designees; bank reviewed on each code revision                   |
+-----------------------------------------------------------------------------+

Re-examination rules belong in the written practice and must be applied uniformly: a stated waiting period before a retake, retraining before a second retake, a cap on attempts, and a fresh examination version each time. For the Level III's own central examinations, ASNT applies its own published waiting periods — a retake after 30 days for a first failure, 90 days after a second, and one year after a third, at which point a new application is required, with no more than three attempts in any two-year period.

Conflict of interest deserves explicit treatment: the person who trained a candidate should not be the sole author, proctor, and grader of that candidate's examination where the organization is large enough to separate the roles. Where it is not, the written practice should require independent review of the grading.