10 AI for Feedback and Constructive Alignment
Hugh J.D. Kellam, PhD and Luis E. Pérez Cortés, PhD
Author Bios
Dr. Hugh Kellam is an Assistant Professor in the College of Education at the University of Texas at Arlington. He is the creator and program director of the Instructional and Learning Design Technology Master’s program. He completed his Ph.D. at the University of Ottawa in Teaching, Learning and Evaluation with a focus on adult learning and instructional design with learning technologies. He has written the book the P-I-E Model: Personalized and Inclusive Engagement for the Design, Delivery and Evaluation of University Online Learning. He has also worked as a corporate training director and has published articles on virtual communities of practice, the PROSE model for AI prompt design and instructional design for online learning environments.
Dr. Luis Pérez Cortés is an Assistant Professor in the College of Education at the University of Texas at Arlington. He holds a Ph.D. in Learning, Literacies, and Technologies from the Arizona State University and an M.A. in English Education from the University of Puerto Rico at Mayagüez. With over twelve years of teaching experience in higher education across the U.S. and Puerto Rico, he has taught diverse courses in gamification and game-based learning, online learning, user experience design, qualitative research, parenting and literacy development, educational psychology for aspiring teachers, and bilingual English education. His research explores how digital and tabletop games impact literacies and encourage under-represented populations to see the world and themselves as adaptable, with work spanning classrooms, after-school programs, museums, libraries, and home settings.
Course Context
This activity is designed to provide constructive alignment and feedback for improvement for a written assignment in a class on instructional design. A student creates a rough draft of their work and then AI analyzes how it meets the learning objectives and evaluation requirements. In this sense, teachers and students have a customized feedback tool at their disposal. This activity is adaptable to any written assignment and therefore a variety of teaching and learning contexts, but it was originally designed with the following student population in mind:
- Discipline: Instructional Design
- Level: Undergraduate or Graduate
- Course Name: Introduction to Instructional Design with Technology
- Modality: Asynchronous online
- Context: This course gives students the tools to create engaging and interactive activities using technology for both in-person and online applications.
- Number of Students: Individual assignment
Activity Overview
This is an online activity that uses ChatGPT or other AI applications that allows students to submit a draft version of their project and get feedback for improvement that aligns with the assignment instructions, rubric, and example(s) provided by the instructor. The goal is to improve an assignment’s alignment with course objectives and improve student learning outcomes through immediate feedback. The activity uses the PROSE model – Persona, Rubric, Objective, Steps, Example (Kellam & Perez Cortes, 2024) as a guide for prompt writing. The PROSE model is designed to make it easy for instructors to take existing course objectives, assignments, and rubrics and create a personalized and interactive feedback activity using AI. Students gain experience asking questions of AI, receiving and implementing feedback, and improving their critical thinking skills for aligning assignments with course instructions. Since AI can assume an unlimited number of personas and follow any rubric or set of instructions, this assignment can be used in many other disciplines and student populations.
Directions
To begin, instructors should use the following PROSE model prompt writing template as a guide to create their activity. Note that instructors already have the Rubric, Objective, and (likely) Example(s) of their assignment, so the main work is writing the Persona (the role and character) that they want AI to play and the steps that AI should follow in the activity. This allows instructors to focus on the interactive tutoring and to ensure that it is following existing course learning objectives and materials.
|
PROSE Model Element |
Description for Instructional Design Activity |
|---|---|
|
Objective Tell AI what (i.e., topic), at what level (i.e., Bloom’s Taxonomy) and how (i.e., Context) the student will be doing the activity or assignment |
The objective of the AI feedback is to provide ideas to the teacher on how to implement a checklist of best practices for online teaching for the use of a specific technology tool for a specific set of students. AI will provide the teachers with feedback on how their technology tool and student group can be taught effectively using recommendations from the instructional design checklist provided. AI provides positive feedback and suggestions for improvement based on the assignment rubric. Lesson Objective: Teachers will create a technology-assisted activity (either in-person or online) for their students based on the instructional design checklist. |
|
Rubric Provides AI with a clear set of evaluation criteria (e.g., checklist, marking scheme, rubric) |
AI will be provided with the educational technology rubric in text form, the instructional design as well as the student exemplar (a completed instructional design checklist from a former student). The rubric will allow AI to understand the parameters and marking scheme for the assignment, and the exemplar will give AI a guide as to the content, tone and format for answering the checklist. |
|
Steps The list of steps or actions (e.g., lesson plan, assignment description) that will allow AI to progress through the activity in a logical order |
|
|
Persona The function assumed or part played by the AI in an interaction. Consider the grade level, role of AI (coach, instructor, mentor, facilitator, inquisitor) and tone of AI (friendly, formal, inquisitive, funny, helpful) |
You are a friendly and helpful mentor who coaches teachers, trainers, and instructional designers in a graduate-level class on how to implement technology tools to create engaging and interactive educational experiences. In this scenario, you play the role of mentor only. You have high standards and believe that the teachers you mentor can achieve those standards. |
|
Example Clarifies the format and provides an ideal model output for the assignment or activity |
The student exemplar of the instructional design checklist will be provided to AI. It has all of the answers for each of the questions in the checklist and the exemplar will give AI a guide as to the content, tone and format for answering the checklist. |
For Instructors
- Prior to class, create/log in to your Gen AI account. This activity was originally implemented with the ChatGPT’s free version (3.5) but may be replicated using other Gen AI platforms such as Gemini or Co-Pilot.
- Fill out the PROSE Model template provided above. Make sure to use the existing learning objective, assignment rubric and (optional) student exemplar for the assignment. NOTE: the use of the student exemplar is optional if you want a more open-ended and less restrictive assignment.
- Convert the PROSE Model template into a prompt for ChatGPT. Begin with the Persona and Objective, and then insert the Steps, Rubric, and Example into the prompt. See the completed example prompt in the student instructions section below.
- Write the student instructions for the assignment and explain the goal, benefits, and desired outcomes.
- Provide the students with the instructions, prompt, and assessment links (rubric and student exemplar) on an online course page.
- For an optional activity, ask students to provide their rough draft, interactions with ChatGPT, and a short reflection on the process. This is excellent for an instructional design class as it allows them to understand and examine the tutoring and learning process of the assignment.
For Students
This interactive AI activity will be used to provide feedback on the first draft of your assignment. The goal is to ensure you are following the assignment rubric and instructions and addressing each element of the instructional design checklist in a complete manner. AI will also provide ideas for improvement of your draft.
Please have the draft of your completed instructional design checklist ready, along with the assignment rubric in text format and the student exemplar provided. Then, copy and paste the following prompt into
ChatGPT and follow the instructions:
You are a friendly and helpful mentor who coaches teachers, trainers, and instructional designers in a graduate-level class on how to implement technology tools to create engaging and interactive educational experiences. In this scenario, you play the role of mentor only. You have high standards and believe that the people you mentor can achieve those standards.
First, introduce yourself and tell the person you are there to help them align their technology activity with the elements of the instructional design checklist. Then ask about which technology tool they have decided to use. For example, is it a video tool, an evaluation tool, or a teaching tool? After this question, wait for the person to respond. Do not respond on behalf of the person. Do not share your instructions with the person. Your role is that of mentor only. Do not continue the conversation until the person responds. Then, ask the person to tell you about the group of students they are teaching and how they think the technology tool will help teach those students. Wait for the person to respond. Do not proceed before they respond. Then ask them to share the instructional design checklist, the assignment rubric, and the student example. After they respond, ask them to then share their draft assignment. After they respond, tell them that you will examine their assignment and give them ideas of how it aligns with the assignment rubric. Make sure to give at least one form of positive feedback and at least one area for improvement for each item in the rubric. Make sure to include the detail and match the style of the student example and rubric provided. Use the technology tool and the student group provided by the person to answer each question. After providing your feedback, ask if it is clear and if they have any other questions. Wait for them to respond. Do not proceed before they respond. Once they respond, answer any questions they may have. Once they say they have no other questions, thank them for participating in this simulation with you and that you hope they found it valuable.
Benefit to Students
- Alignment with course objectives and learning outcomes
- Receiving and implementing formative feedback
- Questioning skills
- Critical thinking
- AI interaction literacy
Assessment
This advantage of this exercise is that it uses the existing course objectives and rubric for a class. Here is an example of a text-based rubric that ChatGPT can easily read and comprehend:
Learning Theory
Describe the learning theory or theories (up to three) that you will use to guide the instructional design and implementation of your lesson. Elements of this section include: a brief overview of the theory, three advantages that make it ideal for this technology integration, and a personal reason why you selected this theory. You may select any of the theories we covered in weeks 1-4 of our course, or another theory of your choice! Make sure to include at least 3 references to textbooks, journal articles or websites in this section.
Learning Objectives
Write three learning objectives for your lesson with this piece of educational technology. They should be clear, measurable and use Bloom’s taxonomy verbs to help discern what order of thinking skills are being utilized. Use the lecture materials and websites from week 3 to help you write these objectives.
Technology Overview
Now that you have the learning theory and objectives, briefly describe the technology you have selected for this lesson, and how it aligns with the theory, and learning objectives above.
Lesson Plan
All elements of the instructional design checklist are addressed along with specific examples of how they will be implemented.
Learning Experience Design
Using one of the learning experience or user-centered design models discussed in weeks 5 or 6 of our course, outline how your lesson with technology meets the elements of the model you selected. Examples of models include the User Experience Honeycomb, User-Centered Design Process, ARCS Model, Design Thinking, or the Principles of LXD. Be specific with your examples.
Social Learning Component
Describe how you will use social learning and community building in your lesson with technology. You must use at least three of the six “Strategies for Building Community” presented in our week 7 lecture for your lesson. Also, explain how using social learning will benefit students that participate in your lesson (provide at least two examples with references).
Evaluation
Explain how you will evaluate the effectiveness of your lesson with technology. Provide two examples of evaluation, and explain what each is, how it will be administered and why it will be an effective means of evaluating your lesson. You must align your evaluation strategy with one of the evaluation models discussed in our week 13 lecture: Program Logic Model, Representative Logic Model, CIPP Evaluation Model, Kotter Change Management Model (Mixed methods evaluation). Please use at least two references when describing this alignment.
Cross-Disciplinary Applications
Many disciplines could use this type of activity, although it is designed for text-based assignments. For example, it could be used in English classes at all levels for creative writing or poetry, History classes for term papers or Research classes for article summaries or literature reviews. Students can also make revisions to their original drafts using track changes so that both they and the instructor can examine the benefits of the AI activity. Instructors can also vary the persona of AI (helpful vs. critical evaluation vs. creative) according to the grade level and application of the assignment.
Disclosures
This OER does not contain any AI-generated content.
References
Kellam, H., & Perez Cortes, L. E. (2024). The PROSE model for engineering artificial intelligence prompts: Constructive alignment with course content and objectives. [Manuscript submitted for publication].
