One more thing, then the module
Day 4, and our Day 1–2 trainer is back in the room after Martin's two days. Before the module proper, she finished a thought left over from the design days — how to stay relevant against the twenty-three-thousand-plus accredited trainers out there. Put 10–20% future content into your programme: what is coming to your industry, taught before anyone else teaches it, because "you got to differentiate yourself from the rest." Close with a teaser, the way Steve Jobs closed keynotes — one more thing — which she filed, correctly, under the law of reciprocity: you were about to leave, and suddenly you stay. And a piece of career surgery: become the sifu of your content and partner for the rest. Content is where your gap is small; technology is where it is wide — "partner with somebody who is good with the technology, because your content will always be you." Then the last content module of the course opened: CT 05, Assess Participants' Competence. The logic connecting it to everything before is one sentence long: you designed the training standard, you wrote the learning outcomes — now the learning outcomes must be measured, and the measurement itself must have standards. What rubric, what scoring method, what passing mark, what evaluation criteria. Friday is this module, applied to us.
Four terms that are not the same word
The module starts with vocabulary, and the trainer treats vocabulary as a safety matter. Assessment is the process of gathering information about whether learning outcomes have been achieved. Measurement assigns the number — 85%, eight out of ten, four of five criteria. A test is one systematic procedure for measuring a sample of behaviour — a sample, never the whole person. And evaluation is the value judgment and decision-making built on all of that information. Assessment looks mainly at the learner; evaluation looks at the whole training. Her discipline about words has a backstory: she develops national standards, and she wanted to be a lawyer before she became an engineer-trainer. Standards work trained her to interrogate every word in a document, because "whatever you put in the document, somebody is going to read it, and the reader needs to interpret it how you want them to interpret it." In training documents especially — "tak boleh ada different interpretation. It's not a novel. You're not doing creative writing." A learning outcome, an assessment item, a marking scheme: technical writing, one meaning each.
Why we assess — the pen and the fake engineer
Why assess at all? She held up a pen. A safety officer, a marketer, the cap supplier and a product designer all assess the same pen — and gather completely different information, because the purpose of the assessment decides what counts as evidence. Same object, four assessments. Why you are assessing determines what you measure. Then the story that carried the whole morning. Years ago, a man was offered a senior technical post — general manager level — using an engineering degree. The certificate was real; it just belonged to his brother, who had a similar name. For a year and a half he survived, though engineers around him grew quietly puzzled by his decisions. The unravelling was pure Malaysian small-world: a customer turned out to be from the same university, same faculty — "from which year? I was there. I never knew you" — because the graduating cohort was tiny and everyone knew everyone. HR pulled the file; it was the brother's certificate. Her point was not schadenfreude, it was the professional lesson she repeats in her industry: "you can have a thousand expert opinions, but one test result with evidence is more powerful." Assessment exists because claims are cheap and evidence is not. The rest of the why-list follows from that: certification — no assessment, no certificate; employment — the form, the referees, the first interview, the second interview are all assessment, because "the more assessment I do, the more information I have, and I can make a better decision to employ you"; and career path — assess where you are now so you can plan the gap to where you want to be.
Assessment FOR, OF and AS learning
Assessment serves three different masters, a distinction usually credited to Lorna Earl's work on classroom assessment. Assessment for learning happens during the training and exists for improvement — quizzes, questioning, observation, practice — "today must be better than yesterday, and I want to know how much you have improved." It is formative, and by nature it repeats. Assessment of learning happens at the end — summative, against the whole syllabus, and its purpose "is not to diagnose; it is to certify." SPM is the national example: one exam, at the end, covering roughly 80% of everything, because a certificate has to stand on broad evidence. Assessment as learning is the learner assessing themselves — athletes are the professionals here, racing their own best times. For assessment as learning she had a story I did not expect: she taught mathematics in a boarding school, and her strategy was to feed her classes almost nothing but problems. Students solved, found the patterns, and built their own methods — assessment as the actual engine of learning, not the measurement after it. The year she and her colleague ran the Form 5 cohort that way, roughly two hundred students averaged a GPA of 1.03 — nearly every student an A1. "That is the power of answering questions." The best advertisement for assessment I have ever heard is a maths class where the assessment was the teaching.
Case Study
Mana Mana's evidence problem
Run the fake-engineer lesson over my own operation and it stings. At Mana Mana, "trained" is a claim, not an evidence-backed judgment. A new front-desk hire is trained when the onboarding session happened and the attendance line is signed. That is exactly the standard of proof that hired a man on his brother's certificate: a document that says something occurred, and zero evidence of competence. We have expert opinions — the supervisor "feels" the new hire is ready — and no test results. The FOR/OF/AS structure is the fix, and it maps cleanly. Assessment for learning: during the first onboarding week, daily five-minute checks — process one mock booking, answer three deposit questions — with the results used to steer the next day's coaching, not to grade anyone. Assessment of learning: at the end of onboarding, a live shift segment assessed against the SOP checklist, competent or not yet competent, and the roster only opens once the evidence says so. Assessment as learning: before the assessor scores them, the hire scores themselves on the same checklist — the gap between the two scores is the most useful conversation of the week, and it trains the self-monitoring habit that athletes live on. And one hiring change, straight from the pen example: decide before the interview what we are assessing for. A front-desk interview currently assesses likeability. The job is composure under simultaneous demands — so the work-sample test should be a staged ten-minute window where the phone rings, a guest waits, and a WhatsApp escalation lands at the same time. Gather the information the decision actually needs.
Reflections
The module's opening hour rearranged something for me. I have always felt assessment as the admin tail of training — the form-filling after the real work. Today framed it as the part that makes training a profession rather than a performance: the difference between "I taught them" and "here is the evidence they learned" is the same difference as between the boy who taught Stripe to whistle and a dog that whistles. Day 1's cartoon was about learning; Day 4 is about proving it. And the boarding-school story quietly settled an argument Day 1 started: the 20:80 lean-learning rule, the retrieval practice from the forgetting-curve post, and assessment-as-learning are the same idea wearing three names. Solving the problems was the learning. A GPA of 1.03 is what it looks like when the assessment and the teaching are one thing.