# Facilitator and assessor guide — EDU-SWE-201

Software Engineering Lecturer Practice | Version 1.0.0 | Public pilot, independent SME review pending

## Preparation

Confirm the audience meets: A relevant computing qualification or equivalent software-development experience; basic Python and Git.
Environment: Python 3.12 or later, Git, a text editor; no third-party Python packages required. OpenCV 4.13 documentation is an extension reading.
Reproduce every executable or editorial fixture before a cohort. Inspect official references and record actual versions. Do not claim this short course replaces employer qualification or experience requirements.

## Delivery plan

Allow 2 hours per foundational module: 15 minutes diagnostic retrieval, 25 minutes explanation and worked example, 55 minutes practice, 15 minutes feedback, 10 minutes evidence/cleanup. The capstone has a separate 4-hour planning budget. Adjust based on a pilot and record observed seat time.

For each topic, ask for a prediction before demonstrating the result. Use the module's unique comparison table or trace as the visual explanation. Give hints in stages: expected versus observed result, relevant rule, relevant location, then worked correction. Require a fresh transfer task after a full solution is shown.

## Role-specific evidence

1. Boundary-test table and corrected booking function
2. Design decision record and operation-count comparison
3. Reproducible image-filter worksheet with border policy
4. A 20-minute image-processing lesson, test suite and moderated marking guide

## Capstone and moderation

Design and teach a 20-minute lesson that takes a 5 by 5 grayscale image through a documented filter. Learners must predict the centre value, run the implementation and diagnose a deliberate border-policy defect. Submit a requirements-to-test matrix, an accessible learner handout and a five-case test suite.

Use the five-criterion rubric in Module 4. Level 3 requires correct, reproducible evidence; level 4 adds justified transfer and fault diagnosis. Reproduce at least one artifact. Record assistance and distinguish independent work. Double-mark a sample and all borderline or contested cases where institutional policy requires; agree the actual sampling policy before formal delivery. Preserve reasons for adjustments and provide an appeal/reassessment route.

Critical condition: Unlicensed source images, private learner data or untested numerical claims require correction before assessment.

## Assessment integrity and privacy

All examples and retrieval answers in this package are public formative materials. They are not a controlled summative test bank. Create any high-stakes variants in an approved private assessment service. No public-page score is an official grade. Confirm assessor authorization, evidence retention and learner support arrangements before a formal cohort. Do not publish learner submissions or identifying records.

## Accessibility and quality review

Review heading structure, keyboard access, zoom/reflow, text alternatives, both themes and meaningful feedback. Native tables provide an exact text representation of each worked example. There is no required video; any future video needs captions and a transcript. Record technical and pedagogical SME findings separately from automated build checks. No ISO, WCAG certification, SCORM or xAPI conformance is claimed.

## Review record

Reviewer role/name: unassigned
Date and environment: to be completed at review
Observed checks and evidence: to be completed at review
Decision: pilot pending independent subject review
Remediation and reassessment: record criterion, new task, evidence and decision