Learning Engineering: #WorkingOutLoud on Learning Science

I’m sharing a new section from the Learning Science work, from the writing with my co-authors Sae Schatz and Geoff Stead . When i shared the last section, on Social Metacognition , i mentioned that we are finding a flow, or a habit, with our writing, and i think that has continued into this piece. We are still exploring, but starting to build off each others ideas well (that being said, it was Sae, who created the backbone of this piece). In terms of overall structure, we are starting to talk with more confidence (although no certainty) about overall structure – for now, this piece is probably heading for the third section: so we start with the broad context (why Learning Science – the evolution of Org learning), move into the insights of the Learning Science, and then into this ‘so what’ piece, of which Learning Engineering with form one of a number of components. We will continue to share our work as we create it, as part of #WorkingOutLoud .

We have looked at individual cognition, at social dynamics, at learning technologies, and at the synergies and integration of?learning ecosystems .

We have considered a broad and evolutionary view of?learning in a modern organisation . And now, in this final section, we will consider how we draw it all together: what can we take away,?what can we ‘do’ with?Learning Science , at speed and scale?

In this context, we will look at ‘Learning Engineering’ as an important emerging professional field.

Here’s it’s official definition from the IEEE Learning Engineering group (penned by Jim Goodell):

• ‘Learning engineering is a process and a practice that applies the learning sciences, using human-centred engineering design methodologies and data-informed decision-making, to support learners and their development.’

In other words, learning engineering is an applied discipline that fuses together (1) learning science, (2) human-centred design, (3) engineering methods, (4) data, and (5) organisational performance principles to achieve goals related to learning.

In one sense we can consider Learning Engineering as both a mindset and a process: an aggregation and assimilation of diverse ideas into a professional practice.

We can also consider Learning Engineering as a discipline, a professional field with shared principles, practices, ethics, and terminology.

Principles of learning engineering include things like this:

• Be data-driven throughout the entire learning lifecycle
• Continuously improve learning activities (like DevOps for learning)
• Incorporate UI/UX throughout all aspects (human-centred design)
• Use learning activities to reach a goal (not solely an end to themselves)
• Use the right tech to optimise goal achievement (whether that tech is donkey-carts or AI)
• Actively intertwine learning science, technology, data, UI/UX, and organisational performance principles.

Perhaps this last point is the most important: to?intertwine our insights?with technology, to draw together knowledge and practice from diverse fields: this all speaks of Learning Engineering as both mindset and an evolution of practice. Something we can learn, if we are willing.

—- —- —-

The?Learning Engineering Toolkit ?offers this story about penicillin that helps draw a distinction between the concepts of science and engineering.?

In 1928, Alexander Fleming discovered penicillin. However, the Nobel Prize–winning scientist and his colleagues never developed the ability to produce the drug at scale. By the 1940s, they could still only cultivate a handful of doses – despite the urgency created by WWI, where nearly a third of military deaths were attributed to infections.?

In 1941, British scientists engaged American engineers to help tackle the problem. (This is an incredibly interesting story that we’re glossing over here.) And by the invasion of Normandy in June 1944, companies were producing 100 billion units of penicillin per month. The death rate of military personnel from bacterial infection fell to less than 1 percent.?

The point of this story is to highlight the different but symbiotic relationship between science and technology. Scientists such as Alexander Fleming discover truths about the world as it is. Engineers create scalable solutions to problems using science as one tool in that endeavour. In a similar way learning science and learning engineering are two sides of a coin: On the one side, learning scientists study how our brains gain knowledge and skill, methods for teaching and learning, and ways to assess progress. On the other side, learning engineers use these finds, combined with engineering methods and systematic principles, to produce reliable and efficient outcomes, often at scale.?

—- —- —-

Learning engineering differs from more traditional instructional design methods, because?those classical approaches tend to be more input focused?(for instance, focused on the decomposition of learning objectives or the development of instructional materials) and?more limited in scope?(for instance, focused primarily on designing and delivering good instructional experiences).

As a counter-example, a learning engineer might improve a training program’s outcomes simply by addressing a faulty wi-fi network in an office space, if that was identified as the main bottleneck preventing improvements to workers’ performance.

When we considered the?broad trends of change in Organisational Learning , from codified to co-created, from central to dispersed, from monocultural to personal, from boundaried to continuous, we saw that context, facilitation, support, all become more important. Learning Engineering is a more holistic, informed, and yet also more personalised practice: personalised in that the Engineer operates in the context of the learner, using design approaches that can focus down to the individual, whilst remaining effective at scale.

It’s an approach that operates at the micro scale of one, at the macro scale of all, and within the mid structure of a?Socially Dynamic Organisation . A learnable art, with the rigour of science. This is why we speak of Learning Engineering as a mindset.

Traditional instructional design puts less emphasis on data-driven processes, and it tends to focus on using?explicit assessments?in learning settings rather than collecting data from a wider range of sources, including clickstream behaviours (that is, someone’s mouse and keyboard clicks), trace data (digital ‘fingerprints’ generated through a person’s interactions, such as dwell-time on an e-book page or a response to a forum post), and implicit estimations (such as using algorithms to evaluate someone’s abilities based on their clickstream or trace data as well as other indicators such as the digital credentials they’ve earned).

And most notably,?traditional instructional design processes are usually linear in practice: A course is designed, developed, and then delivered – and perhaps delivered again and again without much change. In contrast, good learning engineers (like good software developers) never stop improving their learning activities, using data-driven processes to constantly optimise their programmes. (Technology-minded readers can picture Learning Engineering as DevOps for L&D)

In addition to these express differences, there are a few cultural characteristics that generally distinguish the learning engineering professional community. They tend to advocate for an interconnected lifelong learning paradigm (the ‘learning ecosystem ’ concept). To enable that paradigm, they also promote the use of open data/interface standards (in other words a modular open-architecture approach) as well as the use of shared competency frameworks and interoperable credentials. Finally, the community also has a noticeable penchant for sharing, and you’ll frequently find learning engineering resources marked with Creative Commons (or similar) permissive licensing.?

How do you get started with learning engineering? There are several good books and articles, so obviously start there. (Sae recently contributed to the Learning Engineering Toolkit, and she recommends it.) You can also join the IEEE’s Learning Engineering group, called ICICLE. Finally, if you want to begin using learning engineering in a professional context, then here are a few questions to start with:

• Data: When do you collect data during design, development, execution, and continuous improvement processes? What kinds of data do you collect, and what kind of insights do you wish you had (if you had a magic wand, so to speak)? Ideally, L&D data sets should connect to wider organisational performance outcomes. What organisational data – outside of L&D – matter to learning and could these be woven into the L&D remit?
• Continuous improvement: Are you using L&D outcomes data to continuously and specifically improve L&D experiences every time that they’re offered? Are you connecting those data across different courses, subjects, functional units, and goals? It helps to think about digital L&D experiences as an ongoing, and every-optimising process, rather than a static, fixed-in-time event.?
• UI/UX: How can you use human-centred design methods such as conducting frequent user testing throughout the design process? If you can integrate Learning Experience Design or participatory design methods (where future learners sit in with the design team), your final event is much more likely to make an impact
• Reach a Goal: How can L&D representatives gain a seat at the C-suite table when it comes to improving organisational outcomes, rather than remaining an isolated department in your organisation? How robust is your L&D shop’s toolkit? You will increase effectiveness of learning if you can shift L&D focus from churning out curriculum, and rather create leeway to identify and creatively address a broader range of human performance goals?
• Tech: How can you optimise the digital learning experiences in your organisation for maximum reach, and adoption. (this may well imply moving away from traditional page-turning courses, towards information-sharing intranet sites (like confluence, slack). How accessible is learning to your employees? Thinking more broadly than just learner delivery, what other technologies might improve the L&D outcomes of your organisation??
• Intertwined: How might you enable a range of experts to solve larger organisational challenges together (eg: learning scientists, testing experts, courseware developers, software engineers, data scientists, organisational improvement specialists, and subject-matter experts) rather than constraining your scope into a more traditional assembly line within L&D, passing work products from one to the next without much interaction?

As you can see,?the change to a Learning Engineering–powered Organisation is systemic and complex, and much of the change will negate, challenge or break existing power and process. So it’s no small thing.

But, the effectiveness we can achieve through this more holistic, continuous, data-driven approach is not insignificant either.

Further, it’s not necessary to fully achieve all Learning Engineering ideals from day-one.?Begin by fostering a Learning Engineering mindset, taking a big-picture approach to Organisational challenges, using data and human-centred design methods to create learning activities, and approaching L&D as a continuous ‘service’ versus a more product-focused practice.

This is no insignificant thing:?the science gives us insight and engineering disciplines give us tools, but we have to build an appropriate culture to gain the benefit. Of necessity, we may therefore have to consider the socialisation of ideas, the need for prototyping and demonstrations in situ, and the capabilities of the intact team. Part of the journey to a ‘Learning Organisation’, or vibrant Learning Culture, is to unwind the legacy structures and processes of more two dimensional and constrained learning design.

All whilst the ship is sailing.