Place each skill strand by artifact
Objective: Run a short performance diagnostic, score observable criteria, choose skip or repair separately for each strand, and reject one global level label.
I can explain which bridge work I need and show the evidence used for every placement decision.
Build the idea
Placement is multidimensional. One learner may read complex prose well yet need quantitative-language or independent-planning repair. A single beginner/intermediate/advanced label hides that profile and can force unnecessary repetition or leave a high-risk gap untreated.
Score observable performance: source meaning, evidence location, paraphrase fidelity, unit and denominator control, listening reconstruction, interaction repair, task decomposition, deadline monitoring, and transfer. Confidence, accent, age, and time in school are context—not pass/fail criteria.
Choose skip, brief repair, or full strand, then publish the evidence and appeal path. A skip requires an artifact at transfer level, not familiarity. A repair ends only after the same criterion passes in a new academic and workplace task without the diagnostic’s model answer.
Language model
- Learner My English is advanced, so I should skip the bridge.
- Advisor Which artifacts demonstrate every strand?
- Learner Reading and email pass, but my units, paraphrase, and plan do not.
- Advisor Placement?
- Learner Skip demonstrated strands; repair those three and retest in two contexts.
Notice
- Place strands, not identities.
- Use observable task criteria.
- Require unscaffolded transfer before exit.