This article summarize  a practical reference for faculty and staff on integrating Artificial Intelligence (AI) into Outcome-Based Education (OBE).

1. Core Concepts of Outcome-Based Education (OBE) and Constructive Alignment

Modern education has shifted from a content-focused approach to one centered on learning outcomes (OBE). This is achieved through Backward Design and Constructive Alignment, which consists of three key elements:

• Learning Outcomes (LOs): What learners are expected to know and be able to do.
• Learning and Teaching Activities (LTAs): Learning experiences designed to help students achieve the LOs.
• Assessment Tasks (ATs): Assessment methods used to determine whether students have achieved the LOs.

2. Defining Learning Outcomes Using Bloom’s Taxonomy

Learning Outcomes should reflect six levels of cognitive development based on the Revised Bloom’s Taxonomy:

• Remembering: identify, list, define
• Understanding: explain, summarize, illustrate
• Applying: calculate, solve, demonstrate
• Analyzing: analyze, compare, connect
• Evaluating: evaluate, critique, justify
• Creating: design, plan, develop

Alignment of Learning Outcomes

Learning outcomes should be linked across three levels:

Institutional Learning Outcomes (ILOs) → Program Learning Outcomes (PLOs) → Course Learning Outcomes (CLOs)

3. Techniques for Designing Assessment Tools

Formative Assessment

Assessment conducted during the learning process to support improvement and provide feedback, such as short quizzes or reflective journals.

Summative Assessment

Assessment used to evaluate learning at the end of a course, such as final examinations or capstone projects.

Rubric Scoring

Rubrics help assess higher-order learning outcomes while reducing bias. Common formats include:

• Holistic Rubrics: Assess overall performance.
• Analytic Rubrics: Assess performance based on multiple criteria.

4. Applying AI to Assessment Design

Generative AI tools, such as ChatGPT, Claude, and Gemini, can serve as teaching assistants by supporting:

Assessment Item Development

Creating application-based questions, case studies, and plausible distractors for multiple-choice assessments.

Rubric Development through Prompt Engineering

Designing rubrics by clearly specifying:

• Role
• Goal
• Context
• Format

Automated Feedback

Generating preliminary feedback and evaluating open-ended responses according to predefined criteria.

5. Ethics and Safety in AI Use

Data Privacy

Personal student information, including names, student IDs, and examination scores, must never be entered into public AI systems in compliance with PDPA regulations.

AI Hallucination and Human-in-the-Loop

AI may generate convincing but inaccurate information. Instructors must always review and validate AI-generated content before use.

Prompt Templates for Instructors

Template 1: Creating Analysis-Focused Multiple-Choice Questions

“You are a university instructor teaching [Course Name]. Create [Number] four-option multiple-choice questions to assess students’ understanding of [Topic]. The questions should focus on the ‘Analyzing’ level of Bloom’s Taxonomy, include a short case scenario, provide plausible distractors, and include detailed answer explanations.”

Template 2: Creating Analytic Rubrics

“Role: Higher Education Assessment and Evaluation Specialist
Task: Develop an Analytic Rubric for assessing a project in [Course Name] aligned with [Specified CLO]. Present the rubric in HTML table format.”