Learning Engineering: New Profession or Transformational Process?

A Q&A with Ellen Wagner

scientists working in a lab

"Learning is one of the most personal things that people do; engineering provides problem-solving methods to enable learning at scale. How do we resolve this paradox?" —Ellen Wagner

Learning engineering is an emerging methodology that attempts to combine theories and practices from the learning sciences with problem-solving approaches from engineering, ultimately to create valid, reliable, and repeatable solutions that can improve learning at scale.

As such, is learning engineering the next phase of instructional design? Will instructional designers trade in their creative jobs to become "learning engineers"?

Many observers see learning engineering as a new professional field. Others, including some already working in the fields of instructional design and development aren't so sure…

Ellen Wagner is a widely recognized technology strategist, innovator, and advisor to both industry and academia as a founding partner at North Coast EduVisory. Wagner, who notably developed the Predictive Analytics Reporting (PAR) Framework, has worked at the heart of some of the most transformational and influential teaching and learning movements. Here, she guides us in an exploration of learning engineering by reflecting on what she calls the "learning engineering paradox": Can the highly personal activity of learning be transformed into quantifiable outcomes enabled by scaled learning solutions?

Mary Grush: What is learning engineering and what is the learning engineering paradox?

Ellen Wagner: Learning engineering is an emergent practice that seeks to combine the theoretical knowledge about human learning with pragmatic problem-solving methodologies from engineering. The hoped-for outcome is the development of empirically valid solutions for improving learning, at scale. As you can imagine, it is a topic of significant interest among innovation and academic transformation stakeholders.

Current interest in learning engineering has been driven in some measure by experts coming from scientific disciplines outside of education and the social sciences. The emergence of "big data", predictive analytics, machine learning, neural networks, and now generative AI have underscored that social science and education research methods are essential but no longer sufficient to accommodate the kinds of research explorations that are possible using research methods from the hard sciences.

The emergence of "big data", predictive analytics, machine learning, neural networks, and now generative AI have underscored that social science and education research methods are essential but no longer sufficient to accommodate the kinds of research explorations that are possible using research methods from the hard sciences.

Learning engineering presents us with a conceptual paradox: Learning is one of the most personal things that people do; engineering provides problem-solving methods to enable learning at scale. How do we resolve this paradox?

After more than 5 years of actively poking and probing at the construct of learning engineering, I am finding that the greatest value offered by learning engineering may come from thinking of it as a process for applying empirical evidence of learning efficacy: to turn learning evidence into action.

The greatest value offered by learning engineering may come from thinking of it as a process for applying empirical evidence of learning efficacy: to turn learning evidence into action.

It appears that this process of transformation is catalyzed by using design methods, from design practices. Learning engineering connects the results from learning science research to targeted interventions used with specific audiences in specific settings, applying research results in practice. Using iterative design techniques, learning designs are evaluated and improved on their way toward being scaled for broader use in a variety of practice settings. This is where engineering methods ensure technological reliability.

In other words, learning engineering gives us a process for connecting science, solutions, and scale in the service of better learning.

Learning engineering gives us a process for connecting science, solutions, and scale in the service of better learning.

Grush: How have instructional designers reacted to the potential of learning engineering in their professional space?

Wagner: I remember reading Audrey Watters's blog on the topic of The History of the Future of the Learning Engineer, reflecting on MIT's announcement that a whole new category of professional called a "learning engineer" would be taking on roles that sounded suspiciously like instructional design jobs. At the time, she snarkily noted that IDs might not mind having an Ivy League job title elevated from "designer" to "engineer".

IDs haven't really seen learning engineering emerge as the direct competition that was originally feared. These days, they are busy taking care of their stakeholders' instructional, learning, and experience requirements. Instructional designers don't have the luxury of doing much iteration or experimentation in their instructional designs. Learning scientists have a different orientation to their work.

There is a school of thought among some learning scientists suggesting that learning engineering is the practical extension of learning sciences in action. These learning scientists do research, data analysis, and classroom experimentation. They build tools to test assertions. They iterate and formatively evaluate and revise. When they finish their experiment they publish their results.

Meanwhile, IDs continue to support faculty and programs. Instructional designers know they need to do more with data-enabled design in their work. They know they need to up-level analysis skills as an important addition to the ID skillset. Remember, traditionally IDs developed design expertise while working with subject matter experts. But we now have access to ample information and data about our learners from the platforms they use.

Instructional designers know they need to do more with data-enabled design in their work… We now have access to ample information and data about our learners from the platforms they use.

Even so, learning engineering is not necessarily the next generation of instructional design. It is more learning sciences-oriented than that. Thinking of learning engineering as the applied extension of the learning sciences is a tough case to make, ontologically. The learning sciences and engineering appear ontologically independent from one another, which means that combining them, or imagining that one is an extension of the other, is not tenable.

Learning science in practice as learning engineering looks more like a quasi-experiment — not an effort to prototype a promising product or experience, much less a fully scaled solution!

Learning scientists collect data from iterative human design efforts, but may not go on to the next stage of development after the experiment is concluded. Designers and engineers are at different points in the overall learning engineering value chain, with both inclined to focus on developing the next phase of their respective parts of the work.

Grush: How else have researchers in the learning sciences contributed to learning engineering?

Wagner: Learning science researchers are creating what they call learning engineering. It's just that their vision doesn't deal much with either the design or engineering parts of this value proposition.

But who can blame them? They are encouraged by agencies including the Institute of Education Sciences and the NSF, and foundations like Schmidt Futures and others to believe that their version of learning engineering is going to drive the future of U.S. education. They may be surprised to learn that new discoveries are relatively less useful in the design and engineering phases of development, where it is more reliable to use established science in real-life designs and in real-life learning settings.

Grush: What changes might you wish to see in the general perception of learning engineering, that would be helpful in aligning these different interests better?

Wagner: We need to find more reasons to believe that learning engineering offers something of great enough value that it behooves us to figure out how to use it better. And as I've noted, for me the most important benefit that learning engineering presents is that it offers a process for applying research about how human learning works in real-life learning settings.

I like to think of this, as I mentioned earlier, as a process for connecting science, solutions, and scale in the service of better learning:

We start with evidence from learning science; we design solutions that apply research for use on specific outcomes, with specific audiences, in specific settings; and once we figure out what works we engineer those solutions to scale our results.

Once we figure out what works we engineer those solutions to scale our results.

More than anything, learning engineering can give us a transformational process that turns research results into designs that guide people into applying what the research tells us will work best. It makes research actionable. People forget just how hard it is to apply research evidence in practice settings, and just how much testing it takes to ensure that a so-called learning solution actually works. We forget that not every single thing we think of is worthy of scale.

Learning engineering can give us a transformational process that turns research results into designs that guide people into applying what the research tells us will work best. It makes research actionable.

If we are talking about producing learning products at scale, implied in that statement is that someone will be giving someone else money to purchase the solutions being produced. Products will need to be really good for that to happen. Academic colleagues testing assertions of efficacy are likely to find that taking that experimental product to scale is an entirely different undertaking than developing a research tool. It's not for the inexperienced or the faint of heart.

Grush: And so, the learning scientists will remain involved, way down the road?

Wagner: Of course. We need the science driving our collective thinking about better learning to be sharp and relevant.

Actively applying learning science in the design and development of our learning assets, objects, and experiences will improve the quality of the products we produce. It's really as simple as that.

Grush: What role does ontology play in clarifying the perceptions of learning engineering, or even identifying communities of practice?

Wagner: Ontology is a technique for organizing and structuring a body of knowledge. It helps define a common vocabulary for contributors who share information within and across domains. It articulates the essential elements that establish a new discipline's place in the world. An ontology creates shared understanding within a domain, enabling better communication and interoperability. It uses descriptors and relationships among constructs to determine if a prospective new discipline is truly unique and special in the place it holds, or if it is simply a new and different way of talking about something that people already do.

Grush: Can you give one or two examples of groups or individuals whose work could serve as models for learning engineering?

Wagner: To be clear, some of the work going on is quite impressive. In particular I really like Ryan Baker, Ulrich Bosser, and Erica Snow's work on the Learning Engineering Research Framework. Norman Bier and colleagues at Carnegie Mellon University are deeply involved in the Open Learning Initiative, developing free, open tools for educators to use when creating open curricula and content. Danielle McNamara and her colleagues at Arizona State University will be co-hosting the upcoming IEEE ICICLE Learning Engineering Conference July 22-24.

You will notice, however, that these efforts are focused on the learning sciences end of the process model I've described. I continue to be grateful for people like Piotr Mitros, an engineering contributor to the Learning Engineering Google Group, who has voiced concerns about the shifting composition of the learning engineering community, when he poses questions like, "Where are the engineers?" and "Between 2018 and 2021, how did learning engineering turn into ed-tech?" (Mitros, P., 2021).

Fundamentally, my reason for turning to ontological exploration is that the roles played by design and engineering may be as important to the success of learning engineering as is the role played by the learning sciences.

The roles played by design and engineering may be as important to the success of learning engineering as is the role played by the learning sciences.

Grush: Will learning engineering help us see what learning designs are worthy of scale?

Wagner: Learning engineering offers us a process for figuring that out! If we think of learning engineering as a process that can transform research results into learning action there will be evidence to guide that decision-making at each point in the value chain. I want to get people to think of learning engineering as a process for applying research in practice settings, rather than as a professional identity. And by that I mean that learning engineering is a bigger process than what any one person can do on their own.

I want to get people to think of learning engineering as a process for applying research in practice settings, rather than as a professional identity… Learning engineering is a bigger process than what any one person can do on their own.

Grush: And then, by focusing on the process, we won't be endlessly picking over the exact specifications of learning engineering and discussing whether or not it's a profession?

Wagner: No, we won't! We'll be too busy leveraging the outstandingly excellent science from our learning science colleagues, implementing proven design approaches from multiple, relevant schools of learning design thinking, and engineering rock-solid, reliable platforms and apps, at scale.

We can resolve the paradox between learning as a private, personal activity and learning as something that gets broadly scaled by ensuring that we leverage design as a way to catalyze and customize the transformational process — from science, to solutions, to scale — so we can all benefit from the synergies among our collective efforts.

We can resolve the paradox between learning as a private, personal activity and learning as something that gets broadly scaled by ensuring that we leverage design as a way to catalyze and customize the transformational process — from science, to solutions, to scale — so we can all benefit from the synergies among our collective efforts.

[Editor's notes: Image created by AI: Microsoft Image Creator from Designer. Note also that Ellen Wagner will join a high-profile panel at the IEEE ICICLE Learning Engineering Conference July 22-24 co-hosted by Arizona State University.]

Featured

  • stylized illustration of a global AI treaty signing, featuring diverse human figures seated around a round table

    World Leaders Sign First Global AI Treaty

    The United States, the United Kingdom, the European Union, and several other countries have signed "The Framework Convention on Artificial Intelligence, Human Rights, Democracy, and the Rule of Law," the world's first legally binding treaty aimed at regulating the use of artificial intelligence (AI).

  • file folders floating in the clouds, with glowing AI circuitry and data lines intertwined

    OneDrive Update Adds AI Agents, Copilot Interactions

    Microsoft has announced new enterprise capabilities in its OneDrive cloud storage service, many of which leverage the company's Copilot AI technologies.

  • happy woman sitting in front of computer

    Delightful Progress: Kuali's Legacy of Community and Leadership

    CEO Joel Dehlin updates us on Kuali today, and how it has thrived as a software company that succeeds in the tech marketplace while maintaining the community values envisioned in higher education years ago.

  • close-up illustration of a hand signing a legislative document

    California Passes AI Safety Legislation, Awaits Governor's Signature

    California lawmakers have overwhelmingly approved a bill that would impose new restrictions on AI technologies, potentially setting a national precedent for regulating the rapidly evolving field. The legislation, known as S.B. 1047, now heads to Governor Gavin Newsom's desk. He has until the end of September to decide whether to sign it into law.