Building Successful Digital Tools at SOCCCD with Human-Centered Design
A Q&A with Bob Bramucci
"Approached thoughtfully, technology doesn't make things less human — it makes things more human." — Bob Bramucci
For decades, developers at South Orange County Community College District have been creating innovative and useful digital tools that automate tasks to assist faculty and benefit students. They've created myriad tools like My Academic Plan, Sherpa, and Smart Schedule that were not only integrated into mainstream administrative and academic advising processes at SOCCCD, they've inspired further development and the buildout of more comprehensive systems. Here, Bob Bramucci, SOCCCD's vice chancellor of technology, explains how a strategy of human centered design has been key to the success of these tool-building development efforts over time.
Mary Grush: Over the years we've become accustomed to seeing lots of development happening at South Orange County Community College District. And there's always something new — often something that builds on things you've already put in place. What is one of your most interesting projects now? Does it build on work you've already done?
Bob Bramucci: One of the projects we're just beginning, and we're really excited about, is something we're calling Syllabuilder. It's another project with faculty — coming hot on the heels of a couple of successful projects where we've worked with faculty and, I think, have built a degree of trust that will foster more collaboration going forward.
Recently, SOCCCD was given a new accreditation requirement to place all syllabi together in a repository. That's not too much of a challenge, so in the midst of fulfilling that accreditation requirement we were wondering how far we could go beyond that, to offer faculty something new, of real value to them — maybe something that would allow them to get more done for their students in the time they had available.
The idea of Syllabuilder goes beyond simply keeping everything in the same place. It's giving faculty a panoply of tools that they can use to add to or change a syllabus very quickly and dynamically. We wanted them to be able to create a living, breathing document where they can actually do scrum processes that use flexible trial and error to find what works.
Syllabuilder may include several tools — I will just give you an example of one of them. For faculty, Syllabuilder may offer an instructional objectives writing helper. These days, it's very common to have a series of learning objectives, so a textbook chapter often starts by telling students what they are expected to learn in that section. Learning objectives are actually hard to write, so in some cases an instructor may have been searching for just the right words to describe the objective, but may have not succeeded, settling on what is almost placeholder terminology — making the objective unclear or overly generalized. Not only can it confuse the student, this may make it more difficult in the grading process to determine whether or not the student achieved the objective.
We found that the instructional objectives writing process can, for faculty, feel like juggling several balls in the air — keeping so many different things in their heads as they strive to finalize the best language. So the idea of an instructional objectives writing helper is to assist faculty in creating their descriptions with tools like drop-down menus that can help them author objectives. The system can keep — help them remember — their draft language, and make it accessible and easily recalled while they work on the most precise and perfect language for any given learning objective.
It's going to be exciting to see what kinds of help we can give faculty now that we can create tools with functions far beyond the rudimentary checklists of two decades ago.
And to get to your question about building on the past: I had developed something similar to Syllabuilder almost as far back as 20 years ago at Cal State University-Fullerton. It was a project to assist faculty by automating feedback in student writing classes. Back then, we identified a spreadsheet that one of the writing instructors had used to check off common student writing problems from a list of prepared comments that could be combined into a kind of form letter. Following that, more tailored feedback could be provided in newly authored comments at the end. Automating this process allowed faculty to offer much more help to students by using grading time more effectively. But technology is so much better now. It's going to be exciting to see what kinds of help we can give faculty now that we can create tools with functions far beyond the rudimentary checklists of two decades ago.
Grush: What are some of the most successful tools that you and your group developed in the past at SOCCCD? Are your partnerships with faculty and students increasing?
Bramucci: We've developed some successful tools over the years. I'll just mention a few. With My Academic Plan, we're getting close to a million plans that students have made with that tool. (Students may make several plans and then choose one to focus on, so this does not indicate a million student users quite yet. But the utility is clear.) Sherpa, probably the first large-scale "nudge" engine in higher education, has delivered more than 10 million nudges to students. And Smart Schedule is a reimagining of college class scheduling through the eyes of our student design team.
A student design team meets about Smart Schedule for SOCCCD. (Photo courtesy SOCCCD.)
I go into these projects with some pre-conceived ideas, but know that the product will emerge reflecting the interactions that we have with faculty and students. And I realize that teaching and learning have changed from when I was in the classroom. So, the dynamic process of deciding what is needed and then refining it with feedback and in partnership with faculty and, increasingly, students, is an important part of these rewarding development projects.
Grush: You've given examples of some of your development efforts. Is there a more general sense of the directions you're going to have, going forward?
Bramucci: We're always looking to the next thing. Our impact until now has been more on administrative and advising processes. To be able to collaborate more with faculty to improve instruction is the next great opportunity.
Grush: Do faculty now appreciate the value of automation more than they have in the past? Are they more accepting of technology?
Bramucci: Yes, but some of that is the changing cohort. New faculty may have taken classes via online platforms and have used digital tools when they were students themselves. And part of technology acceptance is the rising tide that lifts all boats — the general advancement of technology.
Part of technology acceptance is the rising tide that lifts all boats — the general advancement of technology.
Still, we do have the increasing trust that comes with involving faculty in the trial and error process of development. I think that's what teaches them about what technology and the automation of processes can do. Our agile software development method is so much more flexible and inclusive than the old waterfall method of years ago. Agile is probably at the core of what has engendered that trust.
Grush: Is there an even broader view of successfully incorporating technology in our institutions and in instruction?
Bramucci: Yes, and I think that's human-centered design. It's evident to me that we cannot get where we need to go, to achieve the outcomes that we have set for ourselves in education, simply with human effort alone. We have to have a partnership, if you will, between people and machines. But let's consider that partnership carefully.
We've been working with machines for decades, with the idea that we can achieve more from a system if we pay attention to allocating the functions in that system — allocating the functions to humans that they can do better, and to machines what they can do better. This thought has been with us for probably more than a half century.
But, now is a different time. In an age of artificial intelligence, if you stack up what people do best against what machines do best, more and more, the machine is winning those contests. So, if you stick to the simple idea of giving the tasks that the machine does best to the machines, and those that the human does best to the humans, tasks may be given mostly to machines because they can outperform the human.
Perhaps we can approach all this more safely from another direction: Think not just what humans are good at, but what is rewarding to people.
Think not just what humans are good at, but what is rewarding to people.
For example, when we made the My Academic Plan system, initially some of the advisors were skeptical — was this going to diminish or even eliminate their jobs? In the end, we found that the answer was, profoundly, no. Just the opposite. By automating the "catalog flipping" drudgery, the students and their advisors were able to have much higher-level, deeper discussions about what the students wanted to do with their lives, and perhaps examine where the job market is going as well…. Automation elevated these discussions without detracting from tasks previously done by the advisor. Tasks now done by machines may still be associated with the advisor's roles and responsibilities, but today they are achieved in dramatically different ways.
The idea behind human-centered design of a system is that we are designing for and by humans who are an integral and vital part of that system. Their needs and what is rewarding to them must become part of the overall conversation. It's not just a matter of measuring the speed and efficiency of completing tasks — the machine will win most of those "shootouts." And it's not easy or necessarily obvious either — we work hard to have the conversations that will guide our steps to using technology wisely.
Approached thoughtfully, technology doesn't make things less human — it makes things more human.