CHAPTER 4

ADDIE and Other ISD Models

Theoretical models for the practice of ISD go back to the work of Robert Gagné, Jerrold Kemp, M. David Merrill, and other instructional design notables from the first generation of ISD practitioners from WWII to the early 1970s. These models evolved as part of the formalization of the process of designing curricula and were seen as a way of bringing credibility to the emerging field that we know as ISD. It was reasoned that if there were models for the interpretation of the complexities of relativity and psychiatry, the field of instructional design also needed a platform of scientific methodology to explain and legitimatize its practices. As an illustration of the power of models in general, and ISD specifically, many employers require knowledge of ADDIE or one of many other ISD models as a basic requirement for employment.

You will also often hear instructional designers talk about the fact that they use ADDIE or another ISD model when designing. This usually happens during presentations, or it is written into bid proposals. To an even greater degree, it is now considered a litmus test for a designer to have specific knowledge of and experience with at least one ISD model.

In this chapter, we start our examination of ADDIE and other ISD models by taking a look at what models of any science or practice are designed to accomplish. To first define a model, it is necessary to establish the framework for what a model actually is and what it provides to the study and practice of any field of endeavor.

How Models Evolve

Systemicity is the underlying basis for evolving theoretical models. This is because models at their heart are systems, and each element of a model should show how it relates to the other elements within a defined system. Condillac wrote in 1771 that “a system is nothing other than the disposition of the different parts of some art or science in an order where all these supports each other mutually” (Cardey 2013). It is also likely that any model will contain presumptive and assumptive elements that can be reviewed and examined at length with varying degrees of agreement. At times it seems as if everyone sees something different in a model.

Linguist Yves Gentilhomme (1985) provided an inventory of what he saw as the properties of a system for modeling language. This provides a framework for looking at any model, in our case ISD models. The three basic elements (revised for our use) are:

• Identification: The model must be identifiable as a single object:

   Only one idea

   Able to be named

   Must be distinguishable from other systems

   Must be distinguishable from its environment

• Structure: The model has an internal organization that can be studied

• Interdependence: Each component of a model must have a relationship to the other elements

This systems approach to modeling fits nicely with ISD, since instructional design is considered a systems approach to designing curriculum. Within that framework, each ISD model is a representation of a system for the process of instructional design, and each model is developed using a systems approach to each of a model’s different elements. It is comforting to know that the idea of modeling systems goes back many generations, and the use of modeling in ISD is another indication of the science of the practice of instructional design.

Another term that helps establish the foundational construct of models is mind mapping. A mind map is “a comprehensive, visual and graphical intellectual tool that allows one to organize and better express ideas and creative concepts” (Jablonski 2017). The key words in this definition are visual and graphical. Being visual and graphical not only helps to make an approach look legitimate but also aids in the study and practice of a model.

Most formal definitions of a model will point to the fact that the concept of modeling should provide an “accurate theoretical description in order to understand or explain” if used as a transitive verb (Collins Online English Dictionary) and not as a noun as it is typically used in ISD. It is meant to provide the framework for replicating all or part of a model’s elements. The visual elements of a model represent the relationship of one element to the others in a way that enhances understanding of the way they interact. So, in a hypothetical model, element A is connected to element C, but not to element B except through element D. A picture is worth many chapters of a textbook.

Specifically, ISD models are intended to provide a visual and theoretical road map to define and describe a specific design philosophy and its terms of practice. If a designer follows a specific model and builds workflow activities that reflect the model elements, it is much more efficient to then work through each new design project. It is easy to define each step in the procedure and to determine milestones and evaluation points that will work in most situations. This makes each new project a little less daunting, so the design team can focus on the specifics of each content, population, and implementation choice and not worry about reinventing the ISD wheel with each new assignment.

Common ISD Model Elements

All these models evolved to be vehicles for both explaining and predicting a specific process, in our case the design of instruction. ADDIE very nicely captures the ISD process in a rudimentary, though descriptive manner, and at some level it also is predictive of what is necessary in the process of designing instruction. Some misinterpret the model to suggest that you perform the five elements in a linear pattern, starting with analysis and ending with evaluation. However, a strict conformance to that rendering of the model will usually result in a very simplistic design process with little in the way of dynamic responsiveness to the inevitable complexities and unique aspects of each project.

As we review the five elements of the ADDIE model, be aware that this is only meant to provide a foundation for the design of instruction. It is not meant to be an operational road map to success in the practice of ISD. Which is to say, while we talk about analysis being one of the five building blocks of ISD, we are not prescribing how that will best work within your specific puzzle for each project, with distinct variables in content, population, budget, and other key areas. We will talk about those specifics later in the book. Think of this overview as a look at the guiding principles of ISD rather than the operational steps in the practice of professional instructional design.

To further reflect on the nature of models in ISD, Gustafson (1991) suggested that there is also a taxonomy of instructional design models. It is reasoned that as a science, ISD models needed to classify each operative element and combine them into a system for the design of courses. Given this approach, each model fits into one of three specific focuses:

• Classroom

• Product

• Systems

ISD Models

You can’t have a discussion of ISD models without evoking the names of Robert Mager, Benjamin Bloom, Robert Glaser, Jerome Bruner, Robert Gagné, Leslie Briggs, and Jerrold Kemp; all influenced the evolution of the field. ISD from the late 1960s onward saw a creative and impressive stream of ideas that fostered each of the following approaches, which almost always ended up becoming models.

Among the better known ISD models, and the ones that we will review in this text, are:

• ADDIE

• Rapid Prototyping

• Kemp

• Dick and Carey

• ASSURE

• SAM

Each of these has its unique features and approaches. However, once you break them down into component features, all have the elements of analysis, design, development, implementation, and evaluation somewhere in the mix, regardless of what it might be called in a specific model. As we look at ADDIE and a select group of other models, the wisdom of these models and their learned creators becomes incredibly clear.

ADDIE

The most famous and most commonly discussed ISD model is ADDIE, which is an acronym for the five elements of analysis, design, development, implementation, and evaluation (Figure 4-1). Developed by a team at Florida State University in the 1970s, it was originally designed for use by the U.S. Army but was later adopted by all the military services. It is now universally accepted as the standard used by the majority of designers who claim a preference for one ISD model.

The model was originally called SAT (Systems Approach to Training) and later became ISD (Instructional Systems Development). Over time, ISD went on to be the standard nomenclature for the work generally performed in producing courses, and ADDIE became the name for the theoretical model for ISD.

There are critics of the ADDIE model; that is to be expected when you consider the scope of the ISD practitioner landscape and the need to create niches for new academics and research. Some will argue it is too old and some will argue it is too detailed. It is also considered by some to be too linear and too rigid. Others say it doesn’t work for technology-based learning or online course development. A very small group even suggests that any models for ISD are unnecessary, since it just requires some common sense to design courses.

These discussions are healthy, but frankly there is no other model that does a better job of representing the work that takes place in the practice of ISD at a macro level. Specialist models work very well for some designers and design processes. Designers should review all the models and find the best fit for any specific project.

Figure 4-1. ADDIE

Analysis

Analysis is theoretically the first, or entry, stage of ADDIE, and is considered the data-gathering element for this model. Here, instructional designers assemble all the information they can possibly gather about content, populations, delivery systems, methods, and budgets before they consider anything else. In this model it is assumed that analysis is the first step and that it takes place in each instructional design process, which unfortunately is not always true.

Design

Design is the blueprinting stage of instructional systems, during which instructional designers create the blueprint for a project with all the specifications necessary to complete the project. During this stage, instructional designers write the objectives, construct course content, and complete a design plan.

Development

Materials production and pilot testing are the hallmarks of development. At this stage, most nondesigners not working directly on a project, begin to see progress. Everything from lecture notes to virtual reality is brought from design to deliverable.

Implementation

The most familiar of the elements is implementation. At implementation, the design plan meets the learner and the content is delivered. The evaluation process that most designers and learners are familiar with takes place in this element. Evaluation is used to gauge the degree to which learners meet objectives and facilitators or technologies deliver the project.

Evaluation

Evaluation shouldn’t be listed last in the ADDIE model since it takes place in every element and surrounds the instructional design process. Evaluation is the process of determining both mastery of learners and the quality of a course design. Generally, there are four basic types of evaluation, sometimes erroneously called levels, with many more layers of evaluation possible in any given scenario.

ADDIE Model Tasks

To frame the ADDIE model in operational terms, it is useful to view the key steps in each phase from an instructional designer’s perspective. We will be covering each ADDIE step in greater detail later. Until then, let’s look at the primary functions of each element.

Analysis

• Frame the challenge, problem, or need as tangible action items.

• Determine if each is an instructional or a noninstructional issue.

• Forward noninstructional items to appropriate resources for resolution.

• Evolve strategies for instructional issues.

• Perform the necessary analysis processes to gather data.

• Determine needed resources.

• Draft a budget and timeline.

• Obtain sign-off.

• Evaluate all analysis elements.

Design

• Draft a design plan as your blueprint for the project:

   Rationale

   Objectives

   Population profile

   Course description

   Learner and facilitator prerequisites

   Evaluation strategy

   Deliverables

• Evaluate all design elements.

Development

• Draft the lesson plans:

   Gaining attention

   Recall

   Content

   Application Feedback I

   Application Feedback II

   Application Feedback III

   Evaluation

   Closure

• Draft the materials (where applicable).

• Draft online content (where applicable).

• Pilot test (as applicable).

• Modify as necessary based on pilot-testing evaluation.

• Evaluate all development elements.

Implementation

• Move project to active status.

• Evaluate (Kirkpatrick Levels 1–3):

   Reaction

   Learning

   Behavior

• Modify as necessary based on evaluation.

• Evaluate all implementation elements.

Evaluation

• Review all five ADDIE elements continuously.

• Revise evaluation process as necessary.

Rapid Prototyping Model

As one could assume from the name, the rapid prototyping ISD model “is intended to reduce the time and cost of a traditional ISD approach” (Daugherty et al. 2007). Some think this is the way all design should take place, and it has even been called a “paradigm shift in understanding the nature and purpose of the field of instructional design” (Tripp and Bichelmeyer 1990). This model is based on the fact that certain elements of the ADDIE model and ISD design process are very quickly decided and become boilerplate for future similar designs. Since this model is less linear than ADDIE, it encourages working with already established design decisions in the areas of implementation modes and materials and then building new content on these available modular design elements.

One example of the use of RP is the process of migrating traditional classroom courses to an online environment. While the content of a course will likely remain the same, the delivery methods and packaging of the content will be more streamlined and allow for a level of boilerplating of the design process, which allows for a very productive course build once the prototype is tested for fit and quality. This is incredibly important as thousands of universities and large organizations move their extensive libraries of classroom courses online using one or more learning management systems (LMS). With this approach, the major variables in course design are largely reduced to the content and other course-specific materials. A designer will work with the instructor and any subject matter experts to quickly migrate courses online. This assumes a stable population and an available source of content and materials.

Figure 4-2. Rapid Prototyping Model

Kemp Model

Jerrold Kemp’s ISD model is one of the more interesting approaches to designing curricula. In a world of ISD practitioners who take great comfort in, and occasional criticism for, being linear in thinking and workflow, Kemp’s system is completely circular (Figure 4-3). Unlike in ADDIE or most other models, a designer can start a project from any of the nine elements of this model. It also allows a designer to skip certain elements if they aren’t required for a specific project.

Figure 4-3. Kemp Model

You will find the Kemp model practiced in some language-related course design projects since it is seen to be easily adapted for their specific design needs. This stems from his work at Indiana University and San Jose State University for 30 years in their instructional development programs.

The nine elements are:

• Instructional problems

• Learner characteristics

• Task analysis

• Instructional objectives

• Content sequencing

• Instructional strategies

• Designing the message

• Instructional delivery

• Evaluation instruments

Also included in the model are these elements:

• Planning

• Revision

• Summative evaluation

• Formative evaluation

• Support services

• Project management

This model is a very complex example of the early thoughts on ISD and how a model should represent all the possible process elements necessary to design instruction. The fact that a designer could theoretically start at any of the elements causes some confusion, since an experienced designer would never start with evaluation instruments or content sequencing, for example, before determining objectives and learner characteristics.

Dick and Carey

This ISD model saw wide use during the earlier years of the formal practice of instructional design. Consisting of 10 elements, it roughly follows the ADDIE model. Some see it as “rigid and cumbersome” (Yavuz 2007), and it is certainly a behaviorist design model linking the stimulus of course material with the response of learning the content.

Figure 4-4. Dick and Carey Model

If you look closely at Dick and Carey, you will see all the ADDIE elements with the addition of several specific process areas that would normally be part of any design process but that are highlighted in this model for emphasis. You will see this in many models, including the ASSURE model we will look at next.

ASSURE

The ASSURE model of ISD is the work of Robert Heinich, Michael Molenda, and James Russell. It first appeared in 1989 in their textbook Instructional Media and the New Technologies of Instruction. As the title of the textbook suggests, this model is focused on the inclusion of media in instructional products.

The six elements of the model are:

• Analyze learner

• State objectives

• Select methods, media, and materials

• Utilize media and materials

• Require learner participation

• Evaluate and revise

Figure 4-5. ASSURE

This model is similar to ADDIE, with the addition of the more detailed elements like “utilize media and materials” and “require learner participation,” which are assumed to be done using any ISD model. This model simply makes a point of highlighting the use of media and electronic learning support approaches.

SAM

The Successive Approximation Model of ISD is the result of the work of Michael Allen and was first published in the 2012 book Leaving ADDIE for SAM: An Agile Model for Developing the Best Learning Experiences. Allen suggests that the model “is well suited to small projects.” It is designed to address the four criteria of Allen’s “ideal process model”:

• The process must be iterative

• The process must support collaboration

• The process must be efficient and effective

• The process must be manageable

SAM is divided into three elements, which are the preparation phase, the iterative design phase, and the iterative development phase. Each of these three phases is subdivided into more detailed process elements. While it has three basic elements, two less than ADDIE, if you look at the details of the sub-elements it actually contains all of the ADDIE elements. Again, it is another variation of the original ISD model.

The SAM model is a streamlined design approach that some may find more resonant with their workflow demands, especially in online learning. As with all ISD models, it is up to the designer to decide what works best for a specific design situation.

Figure 4-6. SAM

Comparison of Six ISD Models

As you review Table 4-1 comparing the five generic ISD elements with the six most commonly referenced models, there is an obvious symmetry between each model and the generic construct of instructional design. It is up for interpretation whether rapid prototyping includes the evaluation element and if SAM includes analysis and implementation within their separate frameworks. Assuming that both include analysis and implementation, only worded or interpreted differently, what you will find is that each model has all of the five generic elements, sometimes named something different.

Table 4-1. Comparison of ISD Models

The differences are largely in the actual practice of using a model as a framework for workflow and other design project specifics. Many design elements in these models are more process features than actual model elements. For example, the Kemp model element of instructional strategies would seem to be more of a by-product of the process and not a specific process point. Other examples of this include the elements of writing, modification, and mini-lessons in the rapid prototyping model, which can easily be interpreted as process elements and not model elements.

For instructional design professionals, the practical side of this discussion is that most designers will use the individual model elements in a workflow that best suits each organization, designer, and specific combination of design requirements, including content, population, and instructional strategies. In reality, each designer and design project has its own unique model of ISD for that particular set of requirements.

When it comes to explaining the ISD process to an audience that does not have familiarity with instructional design, such as a client group or those writing a proposal or grant, most designers will use the ADDIE model as a basis. The uninitiated have little, if any, interest in the subtle differences between ADDIE and SAM and only want to know that there is some theoretical construct for the process of designing instruction. Almost everyone has heard of ADDIE, and it just makes things easier to keep it simple for the nondesign community. More experienced and knowledgeable clients and managers may want to dig deeper into a specific project’s approach, and it pays to think about and plan for the inevitable questions concerning design approaches.

The Purpose of ISD Models and Three Simple Truths

For as long as ISD has been a professional endeavor, there have been various theoretical models that attempt to represent the way ISD is structured. From the original ADDIE model to SAM, Kemp, Dick and Carey, ASSURE, rapid prototype, and hundreds of others, these models have assumed an almost mythical status.

In reality, while models play an important role in the education of an instructional designer, their impact on the day-to-day practice of ISD is generally best described as foundational. It is wonderful for the design community to review these, usually as they learn the design process. It allows for the study of the theoretical basis for each model’s approach.

ISD models by their very nature represent ideas, speculation, and hypotheses that are generic, loosely defined frameworks. They present a foundation of theoretical elements reflecting some level of predictive power. To have any chance of being useful in general situations, these ideas must be enacted very close to the prescribed values in a model. Additionally, most models are replicable only within a limited range of variables and, at best, to a modest level of precision.

To use the example of analysis, which is an element of most ISD models in some form or another, we find that the operational reality is often very different from the linchpin status often afforded it in a model. In professional practice it is often not funded adequately, if at all, and is not the primary concern of most designers in the way that most ISD models would suggest. In some situations, analysis takes weeks or months and is funded to a desired level. In many other projects, analysis data is assumed to already exist, and little in the way of resources is allocated. In still other projects, there is simply no analysis process beyond that which takes place when writing objectives or designing a lesson plan. The idea that it must be performed first and is the foundation of all design work is sadly not universally the reality in the practice of ISD.

The popularity of ISD models, and the importance some designers place in them, has its origins in the hope that there exists one universal approach, technique, or procedure that reins in the complexity of instructional design and that provides a logical, manageable, and replicable road map to success in every situation. As you would expect, the silver bullet for ISD doesn’t exist. In fact, one of the primary reasons that experienced instructional designers are in such high demand is their ability to aggregate a very complex instructional design process filled with countless variable elements into a finished product, the path to which is based on a loose interpretation of instructional design models.

The reason for this disconnect between ISD models and the all too practical realities of professional instructional design is based on three simple truths that seasoned designers know exist in the practice of ISD:

• No two design projects are ever the same.

• No one designs in a perfect world.

• One design process never fits all scenarios.

Let’s look at each of these in more detail to explore these ideas.

No Two Design Projects Are Ever the Same

Ask anyone working in the world of instructional design and they will probably say that they have never worked on two projects that were exactly the same. There are always variables, including timing, budget, available resources, and availability of subject matter experts, that make each project distinct. If you attempt to align any project with a model, you will likely never go beyond labeling project tasks with a specific model’s elements. Because I use ADDIE as my model, you will probably be able to find elements of analysis, design, development, and evaluation in various aspects of my work, but ADDIE never really drives the workflow; that is the privileged domain of specific project demands.

No One Designs in a Perfect World

Models assume that every aspect of the design process is bound for perfection and that projects never encounter anything approaching the common realities of modern instructional design. Models lack the flexibility to face the unknown that is waiting just ahead on the schedule. The day-to-day challenges of professional instructional design have never and will never be something that a model can predict. That is unless we have a model with the critical element of “stuff happens.”

No Single Design Process Fits All Scenarios

This final truth is often the Achilles heel of any ISD model. If we believe that our first two truths are accurate, there is little rational possibility that any model could ever promise to reflect a course design process and workflow in anything more than generic terms. It is the skill of an instructional designer that pulls all of the various design elements together regardless of the appearance of working within a template to complete a project. There is no greater folly than the notion that the assembly line mentality works in professional design projects. There are certainly similarities, and a workflow process is always recommended, but the sum of design variables is never the same from project to project.

ISD Models Are Just Models

The very real world of detail and precision in the instructional design process can never be more than loosely reflected in a theoretical model like ADDIE. Models are best used to explain the ISD process to the uninitiated, or to represent our work in a five-minute overview to stakeholders and clients or in an RFP for a project.

Like Einstein’s E=mc² in the field of theoretical physics, the theoretical models in ISD suggest a starting point for the discussion of how this foundational knowledge may lead to practical applications. Just as there is a tension in science between the theoretical and applied science professionals, ISD is in the same position with a tension between the modelists and the practitioners. To be sure, this is a healthy discussion, but most practitioners rely more on their years of experience than the best of theories once they are actually working in the field. Years of trial and error and experience with increasing numbers of variables are the real foundation of the practice of professional instructional design.

While some may wish that there is an all-encompassing ISD model that joins all conceivable design work elements into a simple checklist for success, there exists no realistic mechanism for this to happen beyond the skills of an instructional designer. So, let all of the ISD models simply play a role in your thinking about your work and don’t let them be your day-to-day guide in the sense that everything you do somehow falls within a model’s constraints.

Summary

Every aspect of instructional design rests on a solid foundation of research followed by generations of application and refinement by design professionals. ISD models are one of the most compelling elements of the practice of instructional design, and almost all professional design work starts with the knowledge of these models. The most studied of these models is ADDIE; however, there are literally hundreds of ISD models that have been presented and promoted over the years, each with their unique thoughts on how to practice instructional design. As enlightening as these models appear, the seasoned designer will find that these are only the starting point for the practice of ISD and will modify these design elements to fit a specific project’s needs.

CASE STUDY 1

Your design group has a new client who has never worked with professional instructional designers. The client thinks that designing courses is easy and that it doesn’t take a professional designer to do this work. The client is questioning why someone would need an advanced degree to do this work.

Specifics:

• Small marketing company

• 80 employees in five states

• Need five sales courses for new products

• Previous courses were designed and taught by a senior vice president with no ISD or education experience

• Previous courses were not well received and considered a waste of time by employees

Given what you have learned in this chapter about ISD models and the extensive scientific background of the field of ISD, what will you say to your new client about the value of having professional designers and well-designed courses based on scientific study and practice?

CASE STUDY 2

You are the lead instructional designer for a large biotech firm that designs and offers hundreds of courses every year. All the designers in the department have at least an undergraduate degree in education or instructional design. Several designers have advanced degrees in the field.

The organization has recently promoted a manager in the human resources department to lead your department and to determine if there are any efficiencies that can be gained by designing courses in a more streamlined way, since course design is very expensive given the design team’s education and experience. The new manager has no experience in training or instructional design but mentions ADDIE and SAM as examples of efficient ways to design courses.

How would you explain the relationship between ISD models and the real world of instructional design?

Is there a way to be more efficient and reduce costs by strictly following an ISD model?

CASE STUDY 3

You are a PhD student in a program in the department of education at a large private college. Your research and dissertation are focused on the modernization of educational design approaches and how to address real-world design challenges that exist in the ISD community.

You have decided that there needs to be a new ISD model to address the modern practice of ISD.

What will your model look like and what elements will it contain?