A shared problem: preference without a rationale.

“Make it cleaner” does not identify the problem. “It looks better” does not explain a revision. Both activities start with a reader or user task and ask the learner to connect an observable detail to that purpose.

The feedback lab makes the conversation the practice environment. The visual lab makes the artifact itself available for revision. One develops the explanation offered to another person; the other develops the judgment behind a change.

Move from evidence to action.

Backward design aligns the final decision with the intended skill. An ADDIE planning lens connects the proposed performance gap, objective, activity, and evaluation. Worked examples, guided decisions, and explanatory feedback make the reasoning visible before independent practice changes the context.

The feedback scenario reconverges between decisions so every learner practices all three skills. Signal / Noise accumulates revisions, then resets to a new brief. Both permit exploration while preserving first-attempt evidence.

Use remediation to locate the missing connection.

Vague critique needs a more observable detail. A premature solution needs a clearer connection to the task or constraint. A visual revision needs a rationale beyond taste. Feedback targets those gaps and invites another attempt rather than treating all incorrect choices as the same problem.

Assess beyond the available options.

Selected responses demonstrate recognition within a constrained activity. To examine independent judgment, a proposed evaluation would ask learners to critique and revise an unfamiliar artifact, then explain the relationship between evidence, reader task, and proposed change.

Comparable before-and-after work and a delayed task could be reviewed with a shared rubric. A later workplace feedback sample could provide transfer evidence. Patterns across learners would also inform revisions to the examples, prompts, and feedback.

Original concepts with fictional practice materials. The examples are not client results or student work. Functional verification is distinct from learner testing; no measured learning gain is claimed.

My contribution.

Performance objectives, original practice materials, scenario writing, diagnostic options, explanatory feedback, transfer criteria, visual direction, and iteration, with AI-assisted implementation. The experiences share a learning purpose while keeping their distinct conversation-lab and editorial-review identities.