Start with the performance.

The goal was specific: help reviewers identify observable evidence, connect it to the intended task, explore the designer’s reasoning, and agree on a revision they can check.

That performance goal shaped the experience. Learners step into a design review with Maya, a colleague whose onboarding page needs a clearer first action. They choose what to say, see her response, and examine what their feedback helps resolve or leaves unclear.

Frameworks in practice.

The project uses ADDIE as a planning lens and a SAM-style iterative approach to prototyping. These describe the development approach, rather than a claim that the concept completed a full organizational implementation.

The ADDIE lens connects the proposed performance gap to an objective, practice, assessment, and evaluation plan. Early prototyping made it possible to refine content and layout in context. Review in Wix informed clearer field-guide controls and more consistent alignment of learning objectives.

Backward design connects the assessment to the goal: learners recognize actionable feedback in a new situation, rather than simply recalling steps.

Practice, consequence, transfer.

1. Name the evidence. Replace a general reaction with an observable detail and its likely effect.
2. Explore the reasoning. Work through a constraint without dismissing it or abandoning the goal.
3. Agree on the next step. Define a revision and an observable check.

Options reflect plausible workplace habits: vague critique, premature prescriptions, and avoiding disagreement. Each choice produces an immediate response and an explanation. Learners can explore alternatives while retaining their first-choice record.

The conversation reconnects after each decision so everyone practices all three skills. An independent task then shifts from an onboarding page to an instructional video. A concise field guide supports application beyond the module.

Measure performance, then respond.

Prior functional checks covered 81 complete decision paths, plus retry and restart behavior. Learning efficacy requires a separate check: can learners produce useful feedback themselves?

A formative evaluation would use three measures:

• Before and after: Have learners write feedback for two comparable design problems. Assess both against observable evidence, task connection, inquiry, and an actionable revision with a check.
• Decision reasoning: Ask learners to explain their independent-practice choice to distinguish understanding from selecting the most polished response.
• Delayed application: Review a new feedback sample one to two weeks later to look for retention and transfer.

Remediation would target the missing skill. A reviewer giving vague critique would practice identifying observable details; one prescribing edits without inquiry would practice asking a focused question. Each would receive an annotated example, complete a new attempt, and be reassessed against the same criteria.

Patterns across learners would also guide revisions to the module. If several people miss the same criterion, I would examine the instruction, practice, and assessment before attributing the difficulty to the learners.

These are proposed evaluation methods. Learner outcomes have not yet been measured.

My contribution.

Instructional concept, scenario and assessment design, visual direction, and iteration, with AI-assisted development.