Assessing what students can do
Connect a learning outcome with a task, scoring criteria and evidence of individual understanding.
An assessment supports a claim
A grade is a judgment about performance on particular tasks. An editorial assessment map starts with the intended claim: for example, that a student can distinguish a causal inference from a correlation. The task then supplies unfamiliar evidence and asks for a justified conclusion. Fluent prose alone is insufficient evidence of this specific skill.
Use different tasks for different purposes
A quiz can sample many distinctions. A paper allows extended analysis. A data interpretation task makes reasoning about measures and uncertainty visible. An oral follow-up can probe the basis of a written answer. Choose the combination for the intended outcomes; no single task supplies a complete view of learning. This is editorial guidance.
An original four-criterion rubric
For a short study critique, score four dimensions separately: identification of the design; fidelity to the reported results; justification of the inference; and a meaningful limitation. A simple editorial scale is 0 for absent or incorrect, 1 for partly justified and 2 for accurate and well justified. Provide worked examples before using the scale for grades. It is not an official AP or IB rubric.
Feedback needs another opportunity to act
Ask students to revise a specific inference after feedback, or use a new study that requires the same reasoning. Record what changed. Offer accessible follow-up formats, publish expectations in advance and assess the explanation against clear criteria. This is an original teaching procedure, not a validated diagnostic test.