Marico Flycatcher with lots of nagative space

ADDIE versus AGILE – which approach is the best?

Author: Dr A Erasmus.

We as Instructional (Learning) Designers have been battling with the question “should we use ADDIE or should we go AGILE?” The whole issue of ADDIE versus AGILE has been haunting us. In my opinion we should rather focus on taking the best of both worlds and see how we can integrate the two to get the best results to meet our clients’ requirements. To me, it is not an either/or decision.

The ADDIE instructional design model was developed by the Center for Educational Technology at the Florida State University in the 1970’s for the U.S. Army. Initially the model was based on a sequential and rigid linear progression through five phases: (1) Analysis, (2) Design, (3) Development, (4) Implementation and (5) Evaluation. Soon after its development, the model evolved and was revised and adapted to be more practical as can be seen below. On the graphic on the right it is clear that evaluation is happening throughout the process with a less rigid structure where the phases may also overlap.

From the rigid (1975) to a more flexible (1997) ADDIE


Screen Shot 2016-08-22 at 12.08.58 PM(graphics from History of the ADDIE model):

The ADDIE model works well when everybody involved knows how the end product (learning intervention) should look like, what the learning outcomes are, which content will address the outcomes and how learners should be assessed. It also makes planning easier as large pieces of the design and development effort can be planned upfront, which is also simplifying the project management of such efforts. Large scale instructional design and development projects where the content is matured and stable may benefit from ADDIE style scoping. Using a slight variation on the traditional ADDIE model, one can relatively easily complete each phase quicker and go through a few iterations of each that will enable the incorporation of changes into the project – something that can be time consuming and costly if one uses the pure traditional ADDIE approach.

According to Conrad Gottfredson (seen as the father of the AGILE instructional design approach), it offers an iterative approach to learning design and development by implementing five phases (very much the same as in ADDIE): (1) Align, (2) Get set, (3) Iterate and implement, (4) Leverage and (5) Evaluate. 

Screen Shot 2016-08-22 at 12.10.22 PM

The biggest advantage for me of this approach is that it is “agile” and flexible, giving us the opportunity to design and develop learning content faster and more efficient. By nature, this approach is more learner centered than maybe ADDIE is, and allows for collaboration, rapid prototyping, continuous evaluation and feedback, and revising and improving. Although it is largely based on the strengths of ADDIE, it also breaks the linear nature of ADDIE and allows for a more consultative approach where the client is involved right from the start. It allows us to “fail fast” and this can assist to eliminate the time and effort spent on re-development that is oftentimes necessary when following the rigid ADDIE process alone. lists some benefits (somewhat adapted) of a more agile approach to instructional design. An agile approach:

  • focuses on the learner and his/her engagement with the course material
  • produces better quality learning experiences in a shorter time frame
  • reduces the need for extensive “after-the-time” revisions and updates to course material
  • allows for early and continuous collaboration between stakeholders

In summary, I have only briefly touched on two well-known instructional design models.  I do not believe that one specific model is necessarily better than the other, we should rather combine the learnings and best practices introduced by researchers and instructional design practitioners and devise our own hybrid models that meet our (and our client’s) specific needs – kind of “taking the best from all worlds”.


Leave a Reply

Your email address will not be published. Required fields are marked *