Learning Engineering: Making Education More "Professional"
A Q&A with Ellen Wagner
Learning engineering has taken many forms since the term was coined by Herbert Simon back in the 1960s. Ellen Wagner, who chairs IEEE's ICICLE SIG on Learning Engineering Among the Professions offers some perspective — from Simon's original insight to LE's application and potential today.
"The evolution of ed tech has always demonstrated that as tech platforms get more complex, product teams turn to other disciplines to get the expertise they need." — Ellen Wagner
Mary Grush: Some 50-plus years ago, Herbert Simon (who we remember today as a famed economist and Nobel Prize winner), coined the term "learning engineering" — a term we are hearing a lot these days. What was "learning engineering" in Simon's original context?
Ellen Wagner: Back in 1967 Herb Simon shared a radical vision that colleges and universities could improve their professionalism by increasing the use of scientific methods and business processes in university administration and operation. In today's era of accountability, analytics, "big data", and performance funding, Simon's recommendations sound almost quaint, don't they?
But by increasing the use of scientific methods and business processes, Simon believed it would be possible to improve the returns on investment in college infrastructure and operational management, which in turn would lead to increased efficiency and better outcomes in curricular development, teaching, and ultimately, in student learning. Does this sound more familiar? Maybe even a bit more like performance-based funding, something that is already in place in 34 states?
Among his suggested strategies for making colleges and universities more professional settings for teaching and learning, Simon believed there might be value in providing college presidents with a "learning engineer" [see Simon, "The Job of a College President," p. 77] — an expert professional in the design of learning environments.
As Simon envisioned this role, the learning engineer would be an institutional specialist with several responsibilities related to optimizing university productivity: Specifically, they would be responsible to work collaboratively with faculty to design learning experiences in particular disciplines. They would also be expected to work with administration to improve the design of the broader campus environment to facilitate student learning and faculty improvements. And, they would be expected to introduce new disciplines such as cognitive psychology, along with learning machines and computer assisted instruction — remember, this was 1967 — to various disciplines on campus.
Grush: How does the term "learning engineering" enter into conversations today?
Today, "learning engineering" brings the assistance of scientifically focused, problem solving experts to teaching and learning settings.
Wagner: As learning and performance support technology has continued to evolve, becoming more complex and increasingly more sophisticated, eLearning and digital learning professionals working in education, enterprises, and agencies have recognized their need for new, more sophisticated skill sets. Among these are data visualization, programming and coding, and techniques from learning science and data analytics. Today, "learning engineering" brings the assistance of scientifically focused, problem solving experts to teaching and learning settings. At least that is how an engineer interested in developing learning solutions would describe it.
At the same time, instructional designers and developers, the professionals who in most organizations have been responsible for creating learning experiences for training and development purposes, are finding it increasingly necessary to collaborate with development and engineering partners who can support efforts to innovate and implement interactive, interconnected digital teaching and learning systems.
Learning engineering is not just a bunch of guys sitting around developing lots of new ed tech just for the disruption of it. Even though it is tempting to think of learning engineers as the go-to technical people in an enterprise, their technical skills are part of a holistic skill set that values building, testing, and validating solutions, outcomes, and results.
And the more we see data making its way into our decision making paradigms, and the more we have technology platforms available on campuses and in workplaces to link learners with assets and experiences, the more we need to consider what it takes to ensure that those platforms continue to help us generate the analyses and insights needed to determine whether we are achieving the kinds of returns on our technology investments that we want to believe are there.
The bottom line is, we need to be more focused on measures, numbers, results, and ROI if we want to be sure that all the technologies, and all the people we have using those technologies in their jobs and for learning the things they need to learn, and all the institutional support being used to wrap those interventions around learners and faculty, are not actually part of initiatives that could be a waste of our time, money, and motivation.
Even though it is tempting to think of learning engineers as the go-to technical people in an enterprise, their technical skills are part of a holistic skill set that values building, testing, and validating solutions, outcomes, and results.
Grush: Are current discussions of "learning engineering" defining a new domain, where a new discipline, or perhaps a whole new profession could emerge? Or are we mostly just calling for a new form of professionalism in ed tech and education development in general?
Wagner: I think people who are not necessarily in ed tech or learning design might see learning engineering as a whole new discipline. And I would imagine that if one thinks that educators generally have no tech skills, there might be a sense of relief that the learning engineers are on the way to help.
But for those of us who have already been working in the world of education technology, or learning technology, or eLearning, or online learning, or digital learning, odds are we already know a lot of individuals working in roles that have assumed many of the technological and scientifically oriented responsibilities that some think should belong to learning engineers.
It is probably because so many of us have been working in those roles already that we are greeting this notion of learning engineering as a "new" discipline with some degree of weariness. We may even greet the notion that learning engineers will arrive on the scene (and tell professional educators what to do) with a certain amount of eye rolling.
Grush: Is this merging of engineering values with the education technology and instructional design world a natural, easy process?
Wagner: The distance between ed tech/ISD and engineering programs is deeply rooted in methodological, philosophical, and epistemological differences. Ed tech and ISD come from education. Education is a discipline focused on people, culture, and community; it tends to be more humanistic and less overtly focused on scientific methods. It tends to have strong qualitative attributes. Engineering is rooted in scientific methods to solve problems. It tends toward being more quantitative than qualitative. It is methodologically more focused on products and solutions, not people. It focuses on project management, risk management, evaluation, and operational efficiencies.
For all of this, the greater the expectation on education to be more scientific in how we communicate methodologies and results, the greater the opportunities for learning engineering. In the current view of learning engineering, the methodologies and practices that professional engineers use are applied in key areas like data science, computer science, and learning science. The evolution of ed tech has always demonstrated that as tech platforms get more complex, product teams turn to other disciplines to get the expertise they need. So, why not turn to learning engineering?
Grush: Do you have a good example of an ed tech developer turning to other — scientific — disciplines or to learning engineering methodologies for help?
Wagner: I have many. In the case of our Predictive Analytics Reporting Framework [PAR Framework] platform development, we hired data scientists for data analytics modeling, and we hired system analysts and computer engineers for building out the platform requirements, the data exchange requirements, and so on. We still found we had to be careful around combining data science solutions with engineering infrastructure. It was not as if we could turn directly to a graduate degree program that specialized in training people how to create ed tech platforms, after all!
In the case of several companies I can think of, including Cengage and Learning Objects, there are several staff members whose jobs were specifically set up as learning engineers — intermediaries between product teams and customers, to help them customize purchased platforms into solutions that are uniquely suited to meet the needs of the purchasing institution. Colleagues working for commercial eLearning companies that produce customized programs using the xAPI specification to connect data sources often speak in terms of using learning engineering methods for producing xAPI-enabled solutions.
Another of my favorite examples of learning engineering in action comes from the great work that Drexel University Online is doing with their collections of virtual and augmented reality resources. The Drexel staff that help fit virtual resources into classroom applications engage with the technical staff from companies producing the assets that are part of the Drexel collection, to make sure that the "off the shelf" products are serving the intended pedagogical purposes when being used in class settings. This is more about learning engineering than about engineers, per se, where scientific problem solving methodologies ensure that resource utilization is optimized. Still, making sure that "off the shelf" meets specific curricular needs is definitely a problem to be solved — a good fit for learning engineering.
Certifications, as the low-hanging fruit of the credentialing world might be the right place to start introducing learning engineering curricula into established disciplines.
It's also notable that there are institutions, including Carnegie Mellon University, Harvard University, and Stanford University that have learning engineering programs in place on their campuses. It's interesting that these curricula may not actually be called learning engineering programs; for example, CMU offers an ed tech masters, since the word "learning" in a degree name is still a bit problematic for the College of Engineering. Boston College is about to launch a masters program in learning engineering. And I rather expect we will see a number of institutions launch learning engineering certification programs during the next several semesters. (Certifications, as the low-hanging fruit of the credentialing world might be the right place to start introducing learning engineering curricula into established disciplines — but that's a whole Q&A on its own.)
Grush: What happens if ed tech developers don't seek out that kind of help?
Wagner: Some of the biggest problems that ed tech platform producers and users encounter are when well-meaning education developers — those who don't have much of a sense of how tech systems work together — create a platform, or apps, or whatever, but don't follow the generally accepted "industry" ways for developing or testing or documenting code. Then the rest of us have to spend copious amounts of time creating patches, extensions, and APIs to make sure those special learning platforms continue to work with other ed tech systems.
In other words, people often "hack" their way into creating "unique" education solutions, and then wonder why they don't generalize.
Reflecting back on Simon's original, 1967 criticism of education and his rationale for bringing learning engineering to campus, we see that he wanted to minimize the need for this kind of hacking, and wanted instead to professionalize analytical methodologies for campus problem solving. Using engineering methods to test, evaluate, and troubleshoot the solutions as they are being built — all with the purpose of supporting the development of effective education technologies and learning infrastructures, systems, and organizations — will help contribute to our own professionalism in developing ed tech solutions that will be more likely to play nicely with one another.
Grush: How might a modern version of "learning engineering" be defined, moving forward — whether it's a discipline or profession on its own or the application of professional skills and scientific methodologies in ed tech/ISD or related disciplines?
Wagner: The professions that are likely to be a part of this conversation and the disciplines that want to be included as part of the epistemological foundation will need to do some serious ontological research. We need to have an idea of what learning engineering is — what it allows us to do differently than we can without it — and provide some exemplars in a variety of practices and settings.
On a very practical level, we need to get past posting wish lists of skills as requirements for jobs where learning engineering skills would be desirable. Instead, we need to consider, actually, what some of the fundamental competencies are for building foundational skills and enabling future skill development, across and within education technology, instructional design, learning engineering, or related disciplines.
Using engineering methods to test, evaluate, and troubleshoot the solutions as they are being built will help contribute to our own professionalism in developing ed tech solutions that will be more likely to play nicely with one another.
Grush: It seems like there are many efforts and groups trying to figure out "what learning engineering is" — let's explore that for a minute. How is learning engineering different from instructional design or instructional system design?
Wagner: In the old days, we used to talk about instructional system design as the professional practice that included all dimensions of instructional design (the problem statements, the audience requirements, the performance expectations, the objectives and goals of the learning experience to be created, selection of delivery systems, the quality benchmarks for knowing how to evaluate the overall efficacy of a design, the selection of the right media for distributing learning, and more). Instructional system designers would also worry a lot about integrating learning theory and cognitive theory into their designs. In the ISD model, instructional development tends to function as the product development side of that coin, dealing first with the creation of the media and technology solutions for learning, and then with all aspects of the actual construction of the learning solutions that the designers had conceptualized in the design.
Whenever the complexity of the technology, tools, platforms, programming, and modeling we used in education settings started to push beyond our normal use cases, education developers found themselves turning to experts in other disciplines to get the skills and expertise needed to get the new tools to work. In the 1990s we hired Web developers, multimedia artists, database analysts, and programmers to help us build our tools and platforms. By the 2000s we were turning to system designers, system engineers, programmers and DBAs, and data scientists. Now we look for business analysts, data analysts, data scientists, computer scientists, learning scientists, systems engineers, and anthropologists.
During the past 10 years or so, the increasing dependence that ed tech companies have on learning scientists, data scientists, and computer scientists pushed some companies to create professional positions for learning engineers that combine some of these skill sets to help figure out better ways to navigate product development, project management, product testing, interoperability standards, and specifications. In the world of instructional design and development, the role of the learning engineer is likely to take on many of the more product-specific and technical responsibilities that the instructional developer has held in the past.
Grush: Will learning engineers supersede or supplant instructional designers? Do we need both?
I'm of the opinion that we need both, at least for the next few years as we learn how to transition into doing this work using scientific methods across the whole ecosystem. We need engineers to make sure we keep a solution focus on the platforms that we produce and to make certain the tech works. We need ISDs to make sure people use the technology systems that ed tech companies keep building as intended, and that the technology focuses on solutions that solve problems that really matter to education and training customers.
Remember that learning engineering won't be just for schools and colleges. Learning engineers also have important roles to play in corporate education and training, in military education and training, and in government education and training. Each of these domains has its own very different requirements for solution generation, so we will need to make sure that learning engineers working in these different environments are mindful of the unique problem/solution sets their stakeholders need.
And I think that any and all programs focused on education technology or ISD have a fantastic opportunity to take a look at where they can begin a conversation with learning engineers, that considers future skills for design and development. It may be that we all end up looking more like learning engineers over time as our methods evolve.
Grush: How could learning engineers make life better for instructional designers?
Wagner: Wouldn't it be awesome to have people on a team with the skills to help showcase some of the technological things our platforms can already do? I know that we are talking about an industry — education — that is not necessarily known for the success of its tech rollouts on campus. Just consider LMS training, for example: How many institutions really train their people to use their LMS fully, the way it has been designed to be used? I'm not suggesting engineers do the training, but I am suggesting that there be enablement teams focused on building particular kinds of essential knowledge and skills.
Any and all programs focused on education technology or ISD have a fantastic opportunity to take a look at where they can begin a conversation with learning engineers, that considers future skills for design and development.
The problem is, most institutions don't want to pay the provider company for training, so when it comes to doing new and cool things with tech (e.g. using location data for social science class assignments), most of us are years behind knowing how to get the most out of the technologies we purchase. Who would we turn to with the following questions: Can you help me build an app that works with the LMS? Can you help me create a classroom dashboard? What's the best way to track informal learning? What about linking sensor data to my training data? Could I embed analytics in my workflow?
Wouldn't it be nice to know that we have that kind of capacity on an enablement team?
Grush: Do current engineers perceive the possibility of working at learning engineering differently from people currently working in ISD or in the learning sciences? Is either side thinking they will have to give up too much to work in a field called "learning engineering" — something that's not in their background or training?
Wagner: Maybe not in the learning sciences, but I would say this is certainly happening in ISD and instructional design settings. There is a fundamental disconnect that happens between or among disciplines that are dealing with many of the same issues but doing so from very different perspectives.
It might be tough to get college of engineering groups and college of education groups to merge directly, because college of engineering people will naturally want to drive a conversation about engineering, whereas education people will want to focus on learning. The ideal focus for both, though, would be on a newly minted discipline formed from the merger of learning and engineering. But, are we there yet?
I think the trick, at least for those of us looking to extend our current skills with learning engineering sensibilities, may be to develop essential skills that will help us be better professionals in the domains where we want to be effective.
My engineering colleagues at MIT, Harvard, and Stanford are quite happy to be engineering educators, with the emphasis on engineering and with education colleagues available to further the education mission and education side of the conversation. But they resist the idea of having learning engineering programs coming out of schools of education.
I've observed that education people don't like having their knowledge of pedagogy, design thinking, and assessment made less critical to a central conversation of "engineering". On the other hand, instructional designers like the idea of being called engineers if it gives them more professional gravitas. I think many IDs still chafe from the loss of prestige that has come from having so many job descriptions (for IDs) tending to bundle everything "tech" (from creating Articulate templates to Web design and LMS management) in with program assessment, curriculum development, and performance evaluation and assessment — all into a single position that, whether we like it or not, tends to focus on providing tech services.
I'm personally not so sure I want to be a learning engineer… But I learned a long time ago that I needed certain kinds of learning engineering skills if I wanted to manage an enterprise technology unit or a national learning analytics initiative — just as I learned long ago that I needed certain business skills if I wanted to be a senior administrator of education technology programs.
I think the trick, at least for those of us looking to extend our current skills with learning engineering sensibilities, may be to develop essential skills that will help us be better professionals in the domains where we want to be effective.
I am more likely to hire and manage learning engineers than I am to be a learning engineer these days, but there's still a darn good reason for me to know what I want from my learning engineers, right? And if what we are really talking about is developing better scientific and systematic processes for managing my learning business, then I think we are on the same page.
Grush: Where do we stand now, in creating a strong base for the future and potential of learning engineering?
Wagner: There are several efforts I could point to, but I'll focus here on the one closest to me — which may, in fact, be the strongest. I've certainly appreciated IEEE's leadership in exploring the boundaries and potential of learning engineering. As the leading global association responsible for everything from professional engineering standards to guidelines for professional education, it's been important for IEEE to weigh in.
IEEE's major vehicle for studying learning engineering, ICICLE [the IEEE Industry Connections (IC) Industry Consortium on Learning Engineering], was formed in December 2017 and first convened in January 2018. It has been meeting as an open virtual community for about one year. ICICLE is a 24-month initiative running from December 2017 to December 2019.
ICICLE has organized workgroups around nine distinct SIGs on relevant topics, from curricula for graduate programs and competencies for learning engineers to learning more about artificial intelligence… from design for learning (our learning science group) to examining virtual, mixed, and X-realities… with specific standards and specifications investigated in three of the groups.
Among its stated goals, ICICLE will provide learning engineers with examples of relevant standards so they may understand them and learn how to implement them. It will also hold the IEEE/ICICLE 2019 International Conference on Learning Engineering, now scheduled for May 20-23, 2019, located on the George Mason University-Arlington campus. A proceedings, to be published in December, 2019 will include both conference papers and a white paper introduction to the field of learning engineering. [Check the link above, for program and registration details.]
It's of significant benefit to the discipline of learning engineering that every future event, gathering, or initiative relevant to LE will always relate to the message about "professionalism" behind Herb Simon's original call for learning engineers.
I believe that ICICLE will continue to be a strong home for learning engineering, long after the conclusion of the 2-year ICICLE initiative. Besides ICICLE, of course there have been other collaborative activities surrounding the topic of learning engineering, such as the DevLearn Learning Engineering Summit hosted by the eLearning Guild in October, 2018, and the Learning Engineering Café hosted at OEB this past December. I anticipate and hope such gatherings will continue to spark additional conversation and ongoing collaborations.
It's of significant benefit to the discipline of learning engineering that every future event, gathering, or initiative relevant to LE will always relate to the message about "professionalism" behind Herb Simon's original call for learning engineers. This incredible foundation puts LE directly in the spotlight as an essential new strategy that brings a growing sense of this professionalism to the operation and management of colleges and universities and the development of education. And for that, there is plenty of room for all of us to begin to engage in active conversation and explorations about learning engineering.
