GPTs for Instructional Design | Nicole Ann Hargrove

Design Thinking in Action

Empathize

Through conversations with instructional designers and stakeholders, I uncovered a consistent challenge: Designers we overwhelmed by repetitive tasks like reviewing video scripts for alignment or generating engaging discussion topics for Canvas LMS.

In Figma, I created a Persona and documented Jobs To Be Done to solidify my core problem statement.

Define

How might we help instructional designers at a large university efficiently integrate GenAI into course development?

Ideate

I collaborated with ChatGPT, Microsoft Copilot, and Google Gemini through a simulated Wishing brainstorming session.

The Wishing technique is all about dreaming big by removing constraints and imagining ideal solutions.

Here is a prompt you can use to kick off a Wishing-style brainstorming session with a GPT tool:

"You’ll be my brainstorming partner in a 20-minute simulation using the Wishing technique. Let’s imagine we have zero constraints—no budget limits, no time pressure, and access to any technology imaginable. Our goal is to come up with the most wild, ambitious, and futuristic features a GPT tool could offer to support instructional designers throughout their workflow.

Here’s how we’ll collaborate:

You’ll start by generating 5 big, bold wishes.
I’ll react or build on them.
Then we’ll keep going—riffing, remixing, and pushing ideas further.

Let’s start the simulation—give me your first 5 wishes!"

Prototype

I developed two GPTs using ChatGPT GPT Builder.

The Video Script Reviewer GPT reviews and provides feedback on instructional video scripts, ensuring tone consistency, slide alignment, and adherence to instruction design guidelines.
The GPT Discussion Creator generates scaffolded discussion topics for Canvas LMS based on discussion type (individual, group, crowdsource) based on learning objectives and any uploaded couse content. This GPT is meant to be used by Faculty.

Each GPT was built using prompt engineering found in the Instructions section, examples in the Knowledge base, and has capabilities for Web Search, Canvas, and DALL-E Image Generation.

Below are some mockups of the GPT Discussion Creator I designed in Figma.

Test

After making iterations to my GPT, I did more testing. I used a Microsoft form to gather both quantitate and qualitative feedback.

Majority are Very Likely to recommend the tool.
Majority rated the discussion topics as generated Very Accurately and all rated them to be Very Relevant.
Testers reported Moderate to Significant time saved.
Impact on Teaching ranged from Slightly Positive to Extremely Positive.

I used network-based recruiting to identify my testers. The group consisted of K-12 teachers, university professors, and non-teachers.

Each user had prior experience using GPT tools and basic to intermediate understanding of GenAI and prompt engineering.

Key Insights

The Testers

In Figma, I created a Feedback-Capture Grid to capture feedback, questions, concerns, and improvement suggestions.

I used this insight to iterate on my GPT.

Next Steps

The feedback I received for improvements and challenges has me considering the following:

Usability feedback highlighted the need for a clearer starting point. I am going to add a brief "Start Here" message with step-by-step instructions to guide users before they begin.
Testers reported errors when attempting to download files. Since ChatGPT can sometimes generate phantom files, I plan to mitigate this by offering a copy and paste plain text version of the discussion topic instead.
A tester suggested adding presets to skip steps. While I want to preserve creative flexibility, this might be worth exploring through A/B testing.

Designing GPTs for Instructional Design

A human-centered approach to boosting quality and efficiency with OpenAI ChatGPT GPT Builder

These GPTs are not here to replace human insight. They are here to support it, accelerate it, and make space for creativity in the instructional design.

Overview

Why It Matters

Elevates the quality and efficiency of course design workflows.
Frees up time for higher-order thinking and collaboration.
Promotes ethical, intentional AI use in learning environments.
Sets the stage for scalable, sustainable AI integration in Higher Ed.

As part of my Spring '25 Instructional Design Strategist internship with UNC Digital Lifelong Learning (UNC DLL), I used OpenAI ChatGPT GPT Builder to design two custom GPTs—GPT Discussion Creator and Video Script Reviewer.

These custom GPTs support instructional designers in streamlining their workflows while maintaining content quality and alignment.

This work was grounded in Design Thinking and backed my consultations with the instruction design team to ensure creativity and practicality.