Authentic tasks, not trivia

What this lesson covers

1. 01

   The question nobody at work will ask

   "Which of the following is a benefit of empathy?" can be answered correctly by someone who has never spoken to a customer. At work, nobody asks agents to identify benefits. They ask them to handle a refund exception for an angry customer while the queue grows. An authentic assessment uses that situation.

2. 02

   The fidelity ladder

   Recall: "Define the escalation threshold."

   Recognition in context: "Which of these three cases needs escalation?"

   Decision in a realistic case: A branching call where escalation is one option among plausible ones.

   Performance: A simulated or observed call scored on a rubric.

3. 03

   Rewrite the trivia

   What does PDPL stand for? — A colleague asks you to email this customer list to a marketing agency. What do you do first?" (three options)

   List the five steps of incident reporting. — Here is an incident report a colleague filed. Find the two missing elements.

4. 04

   Validity is an argument, not a property

   Messick and later Kane changed how validity is understood: a test is not valid in itself. Validity is the strength of the argument from a score to a decision. For a module gate, the chain is: the items sample the important job decisions, the scoring is consistent, the score generalises beyond these particular items, and the pass decision is linked to a standard someone can defend. Authentic tasks strengthen the first link. They do not fix the others. A realistic scenario with a confusing scoring rule is still a weak argument.

5. 05

   Climb one item

   Take one question from an assessment your organisation uses today. Rewrite it one rung higher on the fidelity ladder. Then write it at the decision rung. Add both versions to your module artifact.