The successive approximation model (SAM) provides a clear pathway to success—measurable and obtainable milestones for marking completion and targeted moments to reach agreement and consensus. The model is clearly defined and manageable, and yet encourages creativity and experimentation. It consistently reveals the design as it evolves, and it does so in ways that all stakeholders can see and evaluate. It helps all team members communicate with each other, contribute, and collaborate.
SAM is a basic iterative process. It is very effective and well suited to smaller projects, especially when 1) an individual works alone or a small team of individuals works in unison and 2) no specialized skills such as software programming or video production require involvement of others.
The process repeats at least three times, starting and ending with evaluation. The initial evaluation looks at the situation, need, and alternative solutions. After each successive cycle of design and development, the emerging solution is evaluated to determine appropriateness and effectiveness. Information gathered in the initial evaluation is also re-examined for accuracy and sufficiency.
In the first iteration, design is kept to just listing objectives, sketching representative instructional treatments, and proposing methods to measure progress. Development is kept to preparing only representative content for each proposed delivery medium and instructional paradigm. Subsequent iterations will refine the work in terms of breadth, depth, and polish.
1. EVALUATE: Begin with a quick evaluation (analysis) of the situation, need, and goals.
- Who are the learners and what needs to change about their performance?
- What can they do now? Are we sure they can’t already do what we want?
- What is unsatisfactory about current instructional programs, if any exist?
- Where do learners go for help?
- What forms of delivery are available to us?
- How will we know whether the new program is successful?
- What is the budget and schedule for project completion?
- What resources are available, human and otherwise?
- Who is the key decision maker and who will approve deliverables?
2. DESIGN: Quickly, but with thought, prepare a rough design for discussion.
- List and organize obvious goals.
- List behavioral objectives for each.
- List ways learner performance can be appraised.
- Select practical and appropriate delivery media.
- Sketch a few sample designs that appear to fit the situation and could reasonably be expected to achieve the goals. Be as visual as possible. (Sketching is an important activity we will return to in some detail later.)
3. DEVELOP: Prepare prototypes using whatever tools can quickly provide a sense of the design idea in application.
- Select representative content to flesh out some of the sketches. Just enough for understanding. “Tra, la, la…” should appear more often than carefully worded text.
- Stay in sketch mode; nothing fancy here. Prepare bullet points rather than paragraphs; use rough art, snapshots, and homemade video rather than illustrations, professional photos, and commercial video.
- Assemble props that instructors or learners might use to perform activities.
- Focus on prototyping learner activities instead of presentation content.
The iterative process returns to evaluation, one notch up. On the next go around, it’s time to:
1. EVALUATE: Determine the success of the first iteration.
- Was enough known about the situation, need, and goals? If not, it’s time for some additional information gathering and analysis.
- What would and wouldn’t work? Get some learners involved to help you decide. If instructors will deliver the instruction, conduct a mock class. Don’t think it’s too early.
- Where do alternatives need to be explored? Perhaps in this iteration, the team needs to sketch two or more alternative designs addressing the same content to compare.
2. DESIGN: Sketch new alternatives or refine previous ideas. If evaluation determines that the previous cycle should be repeated, it’s important to do it. A cycle that clarifies needs or discards initial ideas is a success and an important step forward, but only if taken advantage of. Repeat Iteration 1 if needed. When it’s time to move on:
- Force a new design. Try creating a design that does not incorporate the design from the first iteration. It may be hard and frustrating at first, but people nearly always find that by imposing the restriction of doing something different, they actually create something better—and something they wouldn’t otherwise have created.
- Identify content that previous design(s) didn’t accommodate well and create initial designs for them.
- Flesh out more thoroughly those ideas retained from previous iterations to be sure the designs can serve as a solid foundation. Representative content should be used, but it’s not yet time to work with the full bulk of it. If there are different types of content, however, representative chunks should be chosen and the design expanded as necessary.
3. DEVELOP: Prototypes need to become more thoroughly representative of the final product.
- Prepare learner materials. Prepare a set of learner materials using a format close to the format under consideration for the final materials.
- Test delivery. Delivery means need to be tested soon, in either the next evaluation or the one in the cycle after that. The type of delivery determines the specific tasks.
- If instructor-led delivery is planned, prepare instructor notes and support materials to a sufficient level that an instructor who has not been involved in the project could deliver a segment of instruction.
- If using e-learning, prepare some interactive segments to test user interface designs and instructional approach.
- If using distance learning, test compatibility of presentation materials and learner responses with the communications system.
Iteration 3 is similar to Iteration 2, although as confidence builds that issues have been properly handled, issues must cease to be re-examined. The iterations become much more focused on development than design.
Additional iterations often seem attractive if not compelling, but if all content areas have been included in the first three, it’s rare that additional iterations would return results worthy of the time and effort. It’s usually much better to put the product in use, get experience with it, and then consider another round of improvements.
Evaluation of SAM
SAM is quite simple, but nevertheless produces excellent products quickly. SAM is easy to manage, because if work ever needed to be stopped unexpectedly or an emergency training need arose, the best product possible under the existing circumstances would be available for immediate use.
Instead of having just a design specification, just storyboards, or just preparation for developing a product (as would be the case with other models), SAM produces something of a usable product after only a couple of quick iterations. Development work begins early in the process. This is done primarily to test ideas and assumptions, but it also provides instruction that is usable in an emergency. Early and continuous availability of a usable product is an outstanding advantage of this iterative approach.
SAM allows design correction early and frequently. It encourages creativity, but verifies ideas early enough to make changes if they fail to verify satisfactorily. It’s a process that serves an individual well, but also facilitates teamwork and collaboration by making ideas tangible and concrete. Traditional processes rely on the approval of specification documents—approvals that are hard to win in a timely manner and yet are often given with divergent understandings of what is being proposed. Successive approximation shuns documentation in favor of prototypes that are much less easily misunderstood.
Reviewing the primary criteria above, this model does nicely here as well:
- support for collaboration
- efficient and effective
There are challenges, of course, with every model. SAM is focused on process and is not a cookbook for instructional design. Instructional design can be difficult and often is. Knowledge of principles, experience, and talent are indispensable. While SAM helps less capable designers do a better job, there are two particular challenges even experienced designers may have trouble with:
1. Refining work too soon. Evaluation in the early cycles may indicate that a deeper understanding of the need is required, that the assumed delivery medium wasn’t a good choice, or that different, more, or fewer people need to be taught. For the process to be nimble and self-correcting, it’s very important that only enough effort is invested in design and development through the early cycles to determine if initial directions are valid and no more. In fact, it’s best to assume that first decisions won’t be valid and early design and development work will be discarded. If too much time and effort is put into early design and development, discarding it will be painful rather than seen as a very important step forward. Poor choices are often adhered to because too much time and effort was invested too early in the process.
2. Perpetual cycling. There’s always a better idea. In fact, as projects progress, the quantity of better ideas tends to rise rather than decrease. Insights and ideas get better and better, while discontent with the chosen direction also increases. Repetition, familiarity, exhaustion, or boredom can lead to making changes simply to refresh the project. Time will eventually run out, and excessive fine-tuning can endanger completion. Implementing just one more great idea can delay rollout. And there’s always one more great idea, just one more. But it would often be better to stop, reserve any remaining time and budget for making improvements later, roll out the product, and gain experience with learners. Most likely, the great ideas that just couldn’t wait will be upstaged by much better ideas gleaned from real application.
SAM is a simple concept. It has something of a “just do it” attitude, favoring action that results in a tangible product to review. While nearly every task prescribed by ADDIE and traditional ISD models is performed in SAM1, they are less onerous and more productive in SAM1. Tasks are short and strive less for perfection. If there is enough to go on, there is enough. There’s no need to write a longer paper, collect more data, have more meetings. If there isn’t enough to go on, the situation is not only discovered quickly, but the need is also clarified in context so that the right amount of effort can be put to it.
SAM is a deceptively simple concept. The apparent simplicity of the model might suggest there is little to it, yet that’s not the case; there’s careful work to be done and pitfalls to avoid. Helpful, however, is that the work does not so much serve the model as the model serves the work—assuring each step is productive and everyone remains aware of where the work is heading.