Outcomes that can actually be assessed

If no task could show it, it is not an outcome.

Learning outcomes are frequently written to sound appropriately elevated: the graduate will appreciate, understand, be aware of, demonstrate insight into. None of these describes something a person could be observed doing, which means no assessment task can establish whether they were achieved.

The practical test is to write, next to each outcome, the task that would demonstrate it. Where that cannot be done, the outcome is a sentiment and should be rewritten as the thing the person will actually be able to do. This exercise routinely reduces the number of outcomes substantially, which is an improvement.

The reverse check matters equally. Every assessment task should map to an outcome, and tasks that map to nothing are examining something the course does not claim to teach. Both directions of that mapping are examined in assessment of an application, and gaps in either direction are found quickly.

Level is the other recurring problem. Outcomes must sit at the level claimed, and applications frequently describe a qualification at one level using outcomes appropriate to a lower one. Comparing against the level descriptors in the applicable framework, line by line, is tedious and catches this before an assessor does.

The other benefit of writing the assessment task beside each outcome is that it forces the volume question early. Twelve outcomes each requiring a substantial assessed task implies an assessment load that may exceed what the delivery hours can accommodate, and discovering that during design is much cheaper than discovering it during the first delivery.