Assessing a student's ability to rationally justify complex policy decisions…

Concept: An assessment format is chosen to match the cognitive demand of the objective being measured. Selected-response formats supply the alternatives and record which one the test-taker marks, so the evidence collected is a choice. Constructed-response formats require the test-taker to generate the response, so the thinking that produced it is written down and can be scored directly against a rubric.

Application: Rationally justifying a complex policy decision is a higher-order objective: the student has to weigh competing considerations, bring in supporting evidence and defend a position. The observable product of that ability is a piece of sustained writing, and it can be examined only if the student is required to produce it. Constructed-response items — short-answer, extended-response and essay tasks — demand exactly that production, and their rubrics award credit for the quality of the reasoning rather than for the position finally taken.

Cross-check against the other formats:

• Objective multiple-choice questions record which supplied alternative was marked. The reasoning behind the mark is never written down, so there is nothing for a rater to judge, and a correct mark can also come from elimination or guessing.

• Criterion-referenced tests describe how scores are interpreted — against a fixed mastery standard instead of against the peer group. That is a referencing framework, not an item format; a criterion-referenced test can itself be assembled from any kind of item.

• Factual recall questions target retrieval of stored information — a date, a term, a definition — which sits at the Remember level of Bloom’s revised taxonomy, a lower cognitive demand than defending a decision.

Result: the format required is constructed-response items.