Data Science for All
In today's rapidly evolving digital landscape, data science & literacy have become essential skills for everyone, and not just a handful of AI specialists.
Project Brief
As part of a statewide initiative, I partnered with a team of administrators and faculty from the Georgia State University College of Arts & Sciences and Perimeter College to develop a new course, DATA 1501: Data Science, to introduce the skills and principles of data science to students beyond STEM. We aimed to make data science accessible to everyone, equipping students with the skills to understand and work with data, make data-driven decisions, and contribute to innovative solutions in their fields. This project-based course gives students the opportunity to be creative while applying data science concepts to real-life scenarios. The future impact of the course is significant. The course is foundational for the new Digital Humanities, Social Science Data Analytics, and Technical and Professional Communication certificate programs, which in just their first semester (Spring 2025) already had 20 students enrolled. The course will be added as a Field of Study option for many bachelor's programs soon as well.
Aligned with GSU Strategic Plan: Student Success 2.0 (Goal 1) and Beyond College To Career (Goals 1, 2, 3)
Tools
D2L Brightspace
FeedbackFruits
zyBooks
ChatGPT & Microsoft Copilot
Google Colab
Jupyter Notebook
HTML
CSS
Qualtrics
My Contributions
Led project team including team of SMEs (administrators & faculty across multiple colleges), multimedia designer, learning technologist
Guided faculty through course mapping (goals, scope, sequence) and content creation and curation
Provided art direction for custom banners and graphics to support narrative-driven project assignments
Developed all new assessments for team based-learning
Designed and built LMS site and integrated learning tools
Ensured course was fully accessible and compliant (WCAG, section 508, UDL)
Provided full service consulting and technical support during pilot semester
Developed and implemented feedback instruments to support iterative design cycle
Who was this for?
The WHO was the most important part of this project. For most of my course design projects, I have a single faculty partner who serves as both the SME and the instructor for the pilot. However, because this impact of this course was expected to be so broad spread (non-discipline specific, taught by faculty in different departments in multiple colleges, supported at the college rather than department level, foundational for multiple certificate programs), I was partnered with a whole committee! I needed to tailor my services for not just one, but 5 partners who each had different priorities, workloads, and travel schedules.
In addition to considering who I was partnering with, WHO we were making the course for was also required a special touch. This was a data science course, but for non-computer science majors. With this course, we sought to open the gates to data science literacy and tools for all Georgia State students to give them an edge in a professional marketplace that is increasingly dependent on data skills.
With the people, the “who,” at the heart of this project, specific opportunities emerged.
Working Together: How could “design by committee” be an asset?
Meeting Students Where They Are: How would the challenge to teach truly complex computer science topics to students who might not even know how to use Excel push us to innovate?
Seeing Their Futures: How could this course be a bridge to students’ future course work and careers?
Working Together
As researchers, each of my partners were comfortable with and excellent at working on their own. Creating something truly collaborative, on the other hand, is hard, so how much they were able to accomplish together was such a feat. They also collaborated with generative AI to jump start their process and anticipate how students might use or misuse AI.
Design by Committee
DATA 1501 Design Team in a virtual meeting
In response to a state-wide initiative, the driving force behind creating this course at GSU was College of Arts and Sciences Dean John C. Medlock. He pulled together a team of like-minded faculty from the Atlanta campus’ School of Computer Science (Berkay Aydin and Kiril Kuzmin) and Perimeter College’s Computer Science (Deepa Muralidhar) and Mathematics (Abu Thomas) Departments. Over the course of a year, they joined forces to design, develop, and deliver a brand-new course. They each contributed, gave feedback, and stepped in for each other when needed. I was so proud of them. Full stop.
But working in such a large group was not without its challenges. We began in summer so I had to work around everyone’s domestic and international travel plans to make we maximized the time we could all be together and we set up for success when we had to collaborate asynchronously. Additionally, when we began, the instructor for the pilot had not yet been identified, so when Kiril was brought on board about 3 months in, I had to catch him up to what we’d already accomplished and work him into our process. I leaned heavily on my project management skills to corral all the moving parts.
Project Plan
Scheduling was tricky for this project. I quickly realized I would only have a short window (May) where I would have all my partners’ full attention before they started peeling off for travel and other obligations, and even though the course wouldn’t launch until spring, my focus would be have to shift to other projects in the fall. Given these restrictions, I leaned heavily on a Backwards Design approach. We dove straight into drafting the 3 major projects because I those tasks would benefit the most from the immediacy of dialogue in meetings, while tasks like deciding on the scope and sequence of textbook chapters could be accomplished through asynchronous feedback.
DATA 1501 Project Plan with 4 phases:
1. “Working Hard” focused on hands-on collaboration; 2. “Working Smart,” focused on asynchronous collaboration; 3. “Just Working,” focused on slow and steady polishing; 4. “Pilot,” focused on the first iteration of the course
Collaborating with AI
I knew I could lose a lot of time if I let this newly formed team waffle about where to start, so I turned to generative AI for some help getting things going on developing the 3 major course projects. By providing the team with sample project instructions generated by ChatGPT, we were able to overcome the paralysis of the “blank page” and get working quickly. Because ChatGPT allowed me to immediately translate what I wanted into the faculty’s disciplinary language (or at least get close enough), I think this method proved superior to providing examples from other disciplines, my own previous work, or empty templates.
Screenshot of ChatGPT chat
I used high level (1-2 sentence) project descriptions from the syllabus and known course context (course title, college level, non-majors, approximate position in course sequence) to engineer a prompt for ChatGPT to create sample projects. I further adjusted the prompts to get ChatGPT to create project instructions using the TILT framework (purpose, tasks, criteria).
I used the projects generated by ChatGPT to create working documents for the team. I was transparent with them about what I'd done, and they were impressed (although not surprised) with the quality ChatGPT's work. We completed 2 of 3 projects in the first month of course development.
Our finished products looked very different than ChatGPT's drafts, but the scaffolding the drafts provided were essential to our speedy success.
This approach was so successful that I gathered feedback from the faculty to create a presentation on our workflow: “Shake ‘n’ Bake and (A)I Helped”
Meeting Students Where They Are
While the course used existing textbooks, we spent many meetings debating and reassessing the scope and sequence of the course and what would best serve the needs of non-computer science students at GSU. We tailored a learning journey that guides students through level-appropriate content, to practice and application, before finally leaving them with bridges to further exploration (“What Next?” pages). Throughout, we sought a balance between building engagement and excitement for the course through an appropriate amount of context setting (why data science matters for everyone, etc.) and diving into content because time is short!
Content Strategy
The team had already decided to adopt textbooks from zyBooks, so I tried to maximize that investment for students. I worked with the vendor to combine chapters from multiple textbooks and with the faculty to create custom content where needed. For example, John drafted a chapter and activity on data ethics based on a a recent training he attended on Responsible Computing Education and then Berkay and later Kiril integrated that work on the zyBooks platform and coded the activities and finally I made sure everything connected to the LMS to ensure seamless student access and grade book synchronization. Since there were so many cooks in the kitchen, my role as project manager was vital for making sure we stayed on track and nothing fell through the cracks. We were able to accomplish much of this task asynchronously using comments, track changes, and detailed meeting summaries that I distributed after every meeting.
Screenshot of “Content Organizer” document with a conversation going on in the comments about what topics to include where
Data in Action
Since this course was aimed at a different student population than these faculty were used to teaching (basically everyone but computer science majors), we knew we needed to make an extra effort to make sure that students saw themselves, and their futures, in this course. One way we tried to connect with students was with constant examples of how people used data skills across a range of careers; we called these features “Data in Action” and we included them in every module overview.
HTML content page in LMS with “Data in Action” blurb
Example of “Data in Action” from
Module 4: Python Basics
Python is a versatile programming language that can be applied to a multitude of tasks. Consider yourself an Operations Analyst at a manufacturing company. You can leverage Python to integrate data from diverse departments like sales and logistics, gaining a comprehensive view of operational processes and performance metrics. This enables you to pinpoint areas for improvement and drive operational efficiency.
You'll also see these skills in action in...
Business Administration
Digital Humanities
Journalism
... and many more!
Seeing Their Futures
Knowing that the students taking this course would come from across disciplines, we were very intentional about providing examples and assessments that used datasets that felt familiar and applicable to their prior experiences and/or future careers. And even though it was more work for the faculty, they created opportunities for student choice by having multiple versions of some assessments. We also embedded resources for students to pursue free badges and certifications that will help them build and communicate their growing skillset.
Relatable Assessments
A first assessment or major activity can make or break a student’s experience of the course. It can be an opportunity for them to really connect with and apply what they are learning in a way that validates their interest, or it can majorly shake their confidence and make them question if they belong in the course. We wanted the former for DATA 1501 students so we looked to storytelling and Universal Design for Learning principles to guide us. For the first project, we provided students with 3 versions which they could choose from and we developed framing scenarios that felt common-place and accessible (securing donations for a food bank, working in real estate, writing a travel feature). Leading with a scenario rather than the dataset primed students to focus on the questions they can answer with data without getting intimidated (or bored) with all the formulas.
What’s Next?
As with any introductory course, DATA 1501 was meant to be a bridge to further experiences. We suspected that while the students enrolling in this course might know they needed data skills, they would likely not know what skills or software would best help them achieve their specific data goals. We also recognized that the pace of the course would not allow for mastery of any particular software, so we wanted to provide supports for students who discovered something useful or interesting and wanted to learn more. We met these needs by including a “What’s Next?” page at the end of every module that linked out to additional resources.
Collaborator Feedback
“Mary Helen served as our learning experience designer for DATA 1501: Introduction to Data Science. She worked closely with our diverse team to create an engaging and highly accessible course, incorporating evidence-based learning strategies, inclusive assessments, and interactive materials that supported students from a wide range of academic and technical backgrounds. Her meticulous approach ensured that the course structure, activities and technologies aligned t.o support student success. More importantly, as a direct result of her work, we observed increased engagement and stronger performance on key learning outcomes. Her expertise, creativity, student-centered approach, and commitment to accessibility were instrumental in the success of DATA 1501.”
“I just wanted to take a moment to say how wonderful it is working with Mary Helen on our DATA 1501 project. She has kept us organized and provided many creative ideas for structuring the course (and our conversations). She has made a daunting task much more achievable – and fun! I look forward to continuing to make progress on the course with our team over the next few months.”
“Today in a meeting attended by both President Blake and Provost Parsons-Pollard, a computer science faculty member mentioned Mary Helen Hoque by name as being ‘incredible to work with and so knowledgeable’.”
“The course content developed and polished by Mary Helen Hoque and implemented in iCollege as a template has been very helpful. It’s the most fully developed course I’ve seen at GSU, and it provides a clear structure that helps both instructors and students stay on track. It ensures that students — especially the most engaged ones — have consistent access to the full schedule, learning objectives, expectations, and additional resources from day one.”
