Gaps in the ADDIE Instructional Design Model

gapI have often written in the past about the strengths of using an elearning model, such as ADDIE, for course design, development, and delivery.

I still happen to believe that ADDIE (or derivatives of this framework) tend to capture the most under the instructional design umbrella, but that’s not to say there aren’t any flaws.

As with any model out there, they should be modified depending on the context of the training material, audience, and client. It is during this manipulation of the model that some of the gaps are exposed.

This doesn’t make the model any less effective (especially if you recognize the shortcoming), you just need to be aware of it.

Putting 100% stock into anything is never a good idea with anything. The better approach is to diversify, and the same can be said when you go through the process of creating a training program.

Some Weaknesses

ADDIE is a strong basis for any training event. There are even other models that have emerged with roots back to ADDIE – it certainly has its place. Still, there are weaknesses. Some of the most common faults, as originally shared by InstructionalDesign.org, include:

  1. Typical processes require unrealistically comprehensive up-front analysis Most teams respond by doing very little at all and fail to access critical elements
  2. Ignores some political realities. Opportunities are misses, vital resources aren’t made available, support is lacking, and targets shift.
  3. Storyboards are ineffective tools for creating, communicating and evaluating design alternatives. Poor designs aren’t recognized as such until too late.
  4. Detailed processes become so set that creativity becomes a nuisance.
  5. No accommodation for dealing with faults or good ideas throughput the process.
  6. Learning programs are designed to meet criteria that are measured (schedule, cost, throughput) and fail to focus on identifying behavioral changes.
  7. Post-tests provide little useful information to assist in improving instruction.

Some of these might not be as apparent in your current elearning development projects as others. For example, I never have had an issue with the last item listed here, especially when using Kirkpatrick four-levels of evaluation.

I also do not agree with the fourth identified weakness. For me, this can only arise if the model isn’t being adjusted properly to fit the  situation. Going by the script can very well limit creativity that falls outside of the model.

In the end, you really don’t need to intentionally pick any model when designing a training program. From my experience, most clients will require you to demonstrate your knowledge of ADDIE, but this doesn’t mean that you have to use it.

Nonetheless, I would argue that even a lack of a traditional model will still resemble one. There isn’t anything wrong with leveraging one of these as a wireframe for your implementation. In fact, I would encourage it.

Categories

About the Author:

Justin Ferriman is the co-founder and CEO of LearnDash, the WordPress LMS trusted by the world's leading organizations, such as the University of Michigan, Digital Marketer, WPEngine, and Infusionsoft. Justin has made a career as an elearning consultant where he has implemented large-scale training programs for Fortune 500 companies. Twitter | LinkedIn

3 Comments
  1. Oh Justin, Justin, Justin…
    Once again you have poked my sore spot. Let me blatantly state that I enjoy reading your posts. You make me think, and sometimes even ponder. And I enjoy an academic debate. So please regard everything that follows as not a personal jab, but an academic response intended to help our community of practices.

    In response to your post, I am tempted to say you failed to do your research (literature review) before expounding such unfounded statements. First you targeted everything on “e-learning.” Why? Isn’t this discussion applicable to all learning?
    1. ADDIE is a construct of what is done in the systematic process of developing effective and efficient training/education solutions. You can refer to it as a “descriptive” model (what is done); but it is incorrect to refer to ADDIE as a proscriptive model (how it’s done). Michael Molenda has conducted extensive research to uncover who actually “invented” it. His conclusion was NOBODY did. The processes required (analysis, design, development, implementation and evaluation) was discussed by many early instruction designers and educational psychologists in the 1960s and 1970s in the context of a systems-oriented scientific process; then, by the early 1980s, everyone was using the term. Many mistook the acronym and related discussions to describe “how” to create instructional material and assumed it was a linear process. (I refer back to the Systems theory and Systems approach to anything). Of course, they were wrong. The processes became known by the combined acronym, and some incorrectly assumed that it was a proscriptive model. Molenda’s (2003) research resulted in his conclusion that
    “the ADDIE Model is merely a colloquial term used to describe a systematic approach to instructional development, virtually synonymous with instructional systems development (ISD).
    2. Be careful my friend. First, all of the pioneers in instructional design and learning theory disagree with your statements. Whose definition of “unrealistically comprehensive up-front analysis” are you citing? Read the literature by Gagne, Molenda, Moeller, Merrill, and Rigeleuth to see how important they feel the up front analysis is to developing the solution. If you don’t know the question, how can you find the answer?
    3. Attempting to attribute “flaws in the ADDIE process” by citing examples of bad ID/ISD practices (resulting from a lack of training/education) reflects a limited literature review.
    4. The use of storyboard is nowhere mentioned in classic ISD literature as a fundamental part of ADDIE or ISD. Some use them, some don’t. Some proscriptive ISD models specifically call for them based on the movie industry’s success.
    5. There are no “detailed processes” in the ADDIE construct. Each detailed process (how to do it) comes from a descriptive model (ASSURE, Agile, etc.)
    6. Your comment about “Learning programs are designed to meet criteria that are measured (schedule, cost, throughput)” should be aimed at program management models, not ISD models. I have only stumbled across one business model touted as (ISD) that tries to define and integrate ID techniques with Prog Mgt practices (Agile, which was created for software development as a radical departure from the waterfall approach).
    7. The comment about failing “to focus on identifying behavioral changes” is a little confusing. If I am focusing on Human Performance Training (HPT), then I must define observable changes, develop learning goals and objectives based on those observable behaviors, and define quantifiable metrics. But not all training is HPT. If I am designing an educational solution, I rarely have observable changes in behavior (Driscoll, 2005).
    8. I will agree that post-tests (by themselves) provide little useful information; however, in conjunction with student feed-back, and other best practices such as a school’s Bureau for Evaluative Studies and Testing (BEST), it becomes vital. HOWEVER, all the ADDIE construct calls out is formative and summative assessment. How to do it us up to the Instructional Designer.

    Again, I continually look forward to your posts, and often find myself nodding in agreement. Sometimes scratching my head and wondering. Occasionally, feeling forced to change or modify an opinion. But always thankful for your inputs.

    References
    Driscoll, M.P. (2005). Psychology of learning for instruction. (3rd ed.). Boston, MA: Routledge.
    Molenda, M. (2003). In search of the elusive ADDIE model. Available at http://www.comp.dit.ie/dgordon/Courses/ILT/ILT0004/InSearchofElusiveADDIE.pdf

  2. U of I

    Coming from academia, I’m in agreement with you on the gaps in ADDIE. It’s not serving our current educational needs. When professors develop courses, they shouldn’t necessarily ‘analyze their audience’ first. They need to analyze new research, and study the job market so that they give students useful, real-world skills and knowledge. A better process could produce what I think is the real value in learning, education that applies and engages students in advancing the global community, and encourages new ways of thinking, new practices, and innovations that solve real-life 21st century problems. Adding the research, relevancy, and real-world components makes it: RRRADDIE. 🙂 Thanks for the great post!

0 Pings & Trackbacks

Leave a Reply

86 Reasons To Choose LearnDash ... Show Me →
+ +