I did not begin my career as an instructional designer. Having spent several years in the classroom as a college- and university-level instructor of philosophy and logic, I stumbled into instructional design by way of educational content development as a learning design author and subject matter expert, combining my classroom experience and subject matter expertise to develop online course materials for introductory-level logic and philosophy students.

This combination of years of classroom experience, content development experience, and an aptitude for using educational technology and content development tools to develop high-quality online course materials led naturally to a career move into instructional design when the opportunity arose. What I did not have, however, was an academic career in instructional design, despite almost a decade of practical, hands-on experience developing highly successful online educational products.

This experience of having created successful online educational products in practice, without the need for traditional instructional design models, led to my having a rather skeptical view about the need for traditional instructional design models, a view that these models emphasize form over content, procedure over pedagogy, and even that they are potentially in conflict with the elegant simplicity of teaching and learning at their best.

All of that said, I thought the time would be right to look into the pros and cons of various traditional instructional design models to see (1) what all the fuss is about, and (2) where they succeed and where they fail as methods for instructional design, educational content development, and educational technology product development.

If you are interested in reading further about traditional instructional design models, the following web page contains an overview of the various traditional instructional design models I consider and evaluate below: Instructional Design Central: Instructional Design Models.

The ADDIE Model

The ADDIE model is a fairly linear product development model consisting of the following phases or stages:

1. Analysis
2. Design
3. Development
4. Implementation
5. Evaluation

Because of its linearity, the ADDIE Model can be considered a type of waterfall development model in which each subsequent phase is begun only after its preceding phase is complete (although the Evaluation phase being placed at the end of the overall design model does not preclude intermediate verification of assumptions and evaluation of effectiveness between phases).

The linear nature of the ADDIE model does have some advantages. For example, the ADDIE model front-loads as much research and analysis into the beginning of the overall design process, which allows developers to verify their assumptions early in the design process while delaying time or resources spent on actual development until those assumptions are verified and the overall development direction is finalized. This can translate to lower development costs early in the overall design process, since no time or resources are yet being spent on hands-on development or implementation. In addition, the Development and Implementation phases ideally would have reduced ambiguity in the ADDIE model, since the individuals performing the actual development or implementation should, in theory, know exactly what to build as a result of the Design phase already having been completed and verified.

Despite the seeming advantages of the ADDIE model described above, the ADDIE model also has several inherent weaknesses. The most serious weakness in the ADDIE model is the radical separation of the Design phase from the Development and Implementation phases. Even if these phases are being completed by the same experienced and well-informed individual or development team, separating the Design phase from the hands-on Development and Implementation phases means that there is a danger of making assumptions about what is possible or ideal during the Design phase that do not bear out in reality during the hands-on Development or Implementation phases.

The Development and Implementation stages are rather late in the overall design process to discover any major design flaw and go back to the proverbial drawing board. In my view, it is much wiser to do actual hands-on development (whether in terms of prototyping, developing some sample content, or even developing a functional early version of a new piece of educational technology) concurrently with the Design phase. This allows you to test the assumptions of a design with actual product or content development tools simultaneously, to more easily verify that the design is practical, possible, and that the remaining implementation will go smoothly.

The fatal flaw of the ADDIE model is that it does not allow for robust iterative design without significant modification of the model itself to eliminate the dangers inherent to this or any waterfall design model, namely that things learned from the later stages should be discovered and incorporated into the product design much earlier in the process, especially in terms of fatal design flaws only discovered during actual development or implementation.

The SAM Model

The SAM model of instructional design attempts to introduce iteration into the overall design process. There are several variants on the SAM model, the summary of the SAM model and its associated SAM model process diagram from Allen Interactions are useful for the purpose of this discussion:

• SAM Process (Allen Interactions)
• SAM Process Diagram (Allen Interactions)

The SAM model combines rapid prototyping and early development/implementation with periodic review and evaluation of the work in progress. This addresses one of the fundamental weaknesses of the ADDIE model, which is the delayed start on actual development and implementation. The SAM model, in contrast, front-loads the early stages of prototyping, development, and implementation but also allows for early iterative changes based on what is learned in the course of early prototyping and development.

So the SAM model has clear advantages over the ADDIE model in terms of its iterative development methodology, but the SAM model also has two weakness that I can identify:

1. The SAM model, at least in the form presented above, still has a fairly strong separation between design/prototyping and development/implementation. The best way to correct for this is to think of prototyping not as a separate phase but something that is done as an early stage of actual development, using the actual tools or environment that will be used to built the finished product.
2. Like all formal methods, the danger exists of the development process being bogged down by the method itself. While the spirit of the SAM model is an improvement over the ADDIE model in terms of iteration, early development, and ongoing review/evaluation, those iterative steps should not interfere with the momentum or progress of the development itself. The iterative feedback loops in the SAM model should be enabling, causing genuinely worthwhile improvements to the overall product or content development, but they should be lightweight enough not be mere barriers or hurdles to overcome, and they should not bog down the overall development process.

Action Mapping

Action mapping, advocated by Cathy Moore, is distinct from the ADDIE model and the SAM model described above in an important way. Although the ADDIE model and the SAM model prescribe the overall development process (whether waterfall in nature as in the ADDIE model or iterative as in the SAM model), the ADDIE and SAM models are neutral in terms of what should be developed and why. Action mapping is intended to be a fun and visual way to determine methodically what content should be developed, based on the overall development goals, and to include only those things that are explicitly helpful in meeting those goals.

Action mapping has a web-like structure with the following components:

1. Measurable goal (placed at the center of the action mapping diagram)
2. Desired behaviors that will reach the goal
3. Realistic practice activities for those behaviors
4. The essential information (and only the essential information) needed for those practice activities

Although, strictly speaking, the web-like structure of an action mapping diagram is not necessary to capture the specifics of these four elements in the design of learning activities, arranging this information in a visual web, with the overall goal at the center, helps to reinforce the idea that everything including in a learning activity or course—from the behaviors to be encouraged, to the practice activities reinforcing those behaviors, to the information needed to support those activities—should be in direct service of the overall goal.

Whereas the ADDIE and SAM models are content agnostic (that is, they do not prescribe the nature of the content or product to be produced), I would characterize action mapping as being workflow agnostic (it does not specify the workflow steps or development phases). An action mapping diagram can itself be a component of the planning or prototyping phase in either a waterfall (ADDIE) development workflow or an iterative (SAM) development workflow.

Conclusion

Each of these traditional instructional design models has its advantages and disadvantages. But the heart of a good product development model, including instructional design models, is to be able to iterate, identify potential problems early in the development process, and maintain a focus on quality (pedagogical quality, usability, or otherwise) at every phase, regardless of which development model is followed, without the workflow itself becoming an additional obstacle to overcome.

In terms of the specific instructional design models covered here, the SAM model has clear advantages over the ADDIE model in terms of iteration, but care should be taken not to let the various feedback loops in the SAM model become barriers to the overall development process. When combined with the Action Mapping method for determining which content or activities are best in service of the overall learning objective, the combination of the SAM Model and Action Mapping can help the development process maintain focus on its overall objective while maintaining quality throughout every phase of development.