#DevLearn-313: A New Agile Model – Leaving ADDIE Behind

Summary: Michael discussed the reasons why we should replace the ADDIE model with Successive Approximation, and how Successive Approximation improves the learning experience.

This session was presented by Michael Allen, Chairman and CEO, Allen Interactions.

Analysis, design, development, implementation, and evaluation – ADDIE – are all important steps in the design of effective e-learning applications or any learning program. While there have been many adaptations of ADDIE, many of them were made before today’s tools, challenges, and opportunities.

More efficient and effective processes are needed and available to produce superior learning experiences in less time. Participants in this session learned about Successive Approximation as a next-generation approach that uses time and other resources to effectively produce more creative and engaging e-learning.

How to select the most effective design and development process for your organization

As we consider what approach might be better, the first question to answer is what kind of product do you wish to produce?

According to Michael, success depends on people doing the right thing at the right time. In Michael’s view, we begin with knowledge but the end goal is always some kind of performance. There are a number of instructional approaches that could, in theory, help us achieve success (listed here in order of effectiveness):

• Tell and test: This is knowledge-based only; it is the same content for different learners where the knowledge is forgotten rapidly after the multiple choice test. In this type of instruction, learners are focused on putting the knowledge required to pass the quiz into short-term memory. This is why it is forgotten so quickly.
• InterACTIVE learning: Interactivity means different things to different people, but here Michael is referring to basic interactivity that involves the learner; this is better than tell and test but, depending on the kind of interactivity, not always successful.
• Distributed practice: This approach spaces out the training over multiple sessions. This dramatically enhances performance.
• Motivation + mentoring: This approach offers a much greater capability for translation of learning into performance improvement. Ideally, e-learning will have both of these characteristics.

The essential components of context, challenge, activity, and feedback (CCAF) to design e-learning applications

Context is the setting of the training that helps learners make sense of the training. Challenge is what people face while they’re learning. The activity is the learning situation in which the learner learns and is job relevant. Feedback is about the consequences of learner actions.

CCAF can be enhanced by connecting with our learners through finding out some things about them. At the same time, we should empower our learner and orchestrate the learning in such a way that the learner is put at risk. It is when the learner has control but faces a challenge that learners are most engaged in the learning.

So then, what do we need in a design and development model to provide for successful implementation of these components? Michael’s says our model should 1) efficiently and effectively produce desired products, 2) be manageable by managers, 3) iterative in nature to reduce waste, and 4) unite the team.

The problem with ADDIE is that it was created as a model many years ago when few knew about instructional design but courses still had to be created quickly. During those earlier years, there were also fewer expectations about what the training should accomplish. ADDIE is also made less effective because of the dependencies that necessarily exist between the different sub-steps in ADDIE. In other words, as Michael put it, making a plan and sticking to it, using a “waterfall” solution like ADDIE, virtually guarantees a sub-optimal solution.

Of course, one suggestion to improve ADDIE is to simply flip the tell and test process and test first. Then, when the learner is stuck, it is at that time that we provide them with the help they need. They are then motivated to learn the content as it helps them address the challenge at hand.

However, Michael advocates a more agile development process. Here, Michael borrows from agile software development processes used in IT departments. In this approach to development, the highest priority is to immediately satisfy the customer through continuous improvement in the software. The approach attempts to deliver a working product as soon as possible, and then refine the product as it is used. Thus, an agile approach provides a competitive advantage by welcoming changing requirements, even late in development. Working software is delivered frequently, with a preference for a shorter timescale. As a result, business people and developers work together daily throughout the project.

Michael referred attendees to the agile manifesto for more information on the general process.

Here are the successive approximation models Michael mentioned in the session:

• Start by evaluating, then, design, develop, then return to evaluation and repeat the process until done. In this successive approximation approach, we throw out the idea as quickly as possible, creating a very rough prototype that works. Learners and SMEs then participate in the evaluation process and add value through new ideas. For example, they might say, “Well, this is how I might address that particular process.” After new ideas are incorporated into the design, learners are given the opportunity to practice and additional ideas are incorporated at that, and later, stages. In the practice round for the learning itself, there is feedback that helps the learner learn, and it is followed by a test round where no feedback is given. In this final stage, the learner achieves the objective of the training without support. As development continues throughout, graphics and look and feel are adjusted to finalize the product.
• This model applies where development has to be separated from design (which is not Michael’s preference). In this approach, there is a preparation phase, which includes information gathering and a project kickoff where a couple of prototypes are built in a couple of days to put the project on target; this also helps identify who the decision makers are who must be involved in the development process. The second phase is an iterative design phase, which includes planning and additional design using a design, prototype and review cycle. Finally, in the iterative development phase, there is a design proof, an alpha, beta and gold release; the cycle here is develop, implement and evaluate.

Michael discusses all of these models in greater detail in his book, SAM: An Agile Model for Developing the Best Learning Experiences.