The Testing Point: The One Decision That Determines Whether Your Exam Is Worth Sitting

Every exam question makes a promise. It promises to measure something that matters. Something that tells you whether this candidate is competent or not.

The testing point is that promise. It is the single sentence that says: this is the professional behaviour we are assessing.

When the testing point is strong, the exam is defensible. When it is weak, nothing else can save it. Not clever scenarios. Not well-crafted distractors. Not rigorous standard setting. A weak testing point produces a question that measures the wrong thing precisely.

The Cascade

A testing point does not exist in isolation. It is the anchor for every decision that follows.

The testing point determines the scenario. If you are testing whether a candidate can interpret monitoring data to detect early compromise, the scenario must include monitoring data that is subtly abnormal. Not dramatically abnormal. Subtly. Because the competence you are testing is the ability to notice what others would miss.

The testing point determines the lead-in. If the testing point is about interpretation, the question must ask for interpretation. Not identification. Not recall. Not “which of the following is true.”

The testing point determines the options. If the testing point defines an axis of discrimination, the distractors must sit on that axis. Each wrong answer represents a different, plausible misinterpretation. Not a random wrong fact. A defensible alternative that a less competent candidate might genuinely choose.

When the testing point is absent or vague, every downstream decision becomes arbitrary. The scenario includes whatever the author remembered. The lead-in asks a generic question. The distractors are whatever came to mind. The question looks professional but tests nothing in particular.

The Recall Trap

There is nothing wrong with testing recall. At undergraduate level, foundational knowledge matters. A first-year student should know basic mechanisms, normal values, core terminology. A recall question is the right tool for that job.

The problem arises when recall questions appear in professional examinations. A graduating candidate, or a practitioner sitting a specialty certification, should not be answering questions that a second-year student with a good memory could pass.

And yet they do. Routinely.

The reason is almost always the same. The testing point was never defined with enough precision to prevent it. “The candidate should understand drug metabolism” allows a question writer to ask about a single enzyme. That is recall. “The candidate should predict how concurrent disease alters drug metabolism and adjust dosing accordingly” requires integration, reasoning, and professional judgment. That is professional competence.

The difference between those two testing points is the difference between trivia and assessment. The question writer may have intended the second. Without a precise testing point, they will almost always produce the first.

The Template Problem

Some institutions have tried to solve this with standardised formats. A prescribed structure that looks something like:

“Given a [situation] that presents [data], a candidate will be able to [verb] the correct [content] and select it from a list of options.”

This describes the mechanics of answering a multiple-choice question. It does not describe a competence.

Fill in the blanks and you produce a sentence that sounds rigorous. “Given a patient scenario that presents clinical findings, a candidate will be able to identify the correct diagnosis and select it from a list of options.” What does that tell you? Nothing you did not already know. It is a description of how MCQs work, not a statement of what this specific question must measure.

A useful testing point describes professional behaviour without referencing the question format at all. “Differentiate between two conditions that present with overlapping initial findings, using data that only becomes discriminating when interpreted together.” That tells the question writer exactly what to build. The template version does not.

Templates create the illusion of rigour. They produce compliance without competence. The field has been completed. The testing point says nothing.

Where Platforms Get It Wrong

Open any major question authoring platform. Find the testing point field. It will be at the bottom. After the scenario. After the lead-in. After the options. A metadata field to be completed before submission.

This design tells the author that the testing point is administrative. Something the system requires, not something the question requires. An afterthought. Documentation rather than architecture.

When the testing point lives at the bottom, it becomes a retrospective description of whatever the author happened to write. Not a deliberate decision that shaped the question from the start.

CrtQ puts the testing point at the top. The first field. The banner that stays visible while every other element is built beneath it. The author writes the testing point before anything else exists. The question grows downward from that anchor.

The platform checks whether the testing point is specific and measurable. It flags vague language. It identifies knowledge verbs that signal recall rather than reasoning. And when the question is complete, the author can ask: “What am I testing?”

The AI reads the complete question and reflects back what it actually measures. Not what the author intended. What the question, as written, will assess.

Sometimes those match. Often they don’t. The testing point says “interpret data to detect early compromise.” The question, as written, can be answered by knowing a single normal value. The platform shows the gap. The author sees it. They rewrite.

That reflection is where assessment literacy grows. In the moment. At the point of writing. Not in a workshop six months earlier.

The One Question That Matters

Before you write a scenario. Before you craft a lead-in. Before you think about distractors. Ask one question:

What must a competent candidate be able to do?

Not know. Do.

Write that down. Make it specific. Make it describe a professional behaviour that cannot be answered by recall alone. Then build the question around it.

If the testing point is strong, the question will be strong. If the testing point is weak, nothing you write afterward will fix it.

The testing point is not documentation. It is the foundation. Everything else is built on top of it. And if the foundation is weak, the whole structure falls.

Sharper questions start with sharper testing points. Everything else follows.

CrtQ. Sharper questions. Smarter exams. crtq.ai