How do you set up a progression with backward chaining?
Suppose a task to be performed consists of these steps:
Start ? Step 1 ? Step 2 ? Output
At the risk of repeating myself too much, look again at the options for teaching this task in terms of the point at which the learner produces output. The conventional approach, or forward chaining, is to teach the steps in the order in which they are done on the job. The learner only produces the task output in the last exercise.
In backward chaining, the exercises are ordered so that learners produce output right away. For example, the learner is presented with a nearly completed task, such as a letter on screen that is already typed, spell-checked, edited, and formatted. The learner is immediately shown the steps needed to print the letter, finishing the job. The next exercise will demonstrate how to format text that has been typed, spell-checked, and edited, and the learner will be prompted or coached through the steps needed to print it. The third exercise demonstrates editing, prompts formatting, and releases the learner to print the text without further help. These steps are repeated until the learner can do the whole task unassisted.
Overview of steps
Because backward chaining always seems somewhat counter-intuitive to designers the first time they see it, the specific design process involved may also feel a little “un-natural.” Fortunately, there are only four steps in the process, and they are straightforward and simple.
Before starting the development, as always, you need to have decided what performance will be taught (in other words, you need to have decided on your objectives). You also need to have decided who will be taught (the target audience), and how the instruction could be presented. This last decision involves thinking about synchronous delivery vs. asynchronous delivery and the possibility of blended learning. It also involves thinking about the appropriate use of multimedia and text.
Describe the job. This can take a little time, but it is important to do it right. The success of any e-Learning application rests on the quality of the description and analysis of the task or skill being taught. Remember that countless studies have shown that, when determining success, the medium of delivery (e-Learning, classroom, plain text) matters less than the quality of design.
In this step, you are going to identify, organize, and describe in measurable terms all conceivable situations that signal the start of, or that arise from, the performance, and all mastery level actions that lead to the desired results or goals. As you will see, the result may look as simple as Figure 1 did.
In doing the task analysis, begin with the fewest number of steps. Dividing a skill into smaller and smaller steps moves the learner further and further from the purpose of the activity. If there are steps that turn out to present problems, you can always break them down further later on.
Task analysis must include both steps that can be seen (motor skills) and those that cannot be easily observed (thinking, analysis, judgment). Test the scope of the task analysis by asking if the learner will be able to complete the task in a reasonable amount of time, and will the learner know:
- When to begin the task
- How to prepare for the task
- How to determine if the task has been done well
- What to do if a problem comes up
- What to do at the end of the task
If the answer to any of these is “no,” then the task analysis needs further work before you continue with the design of the instruction.
Group the tasks. Organize the tasks into spans of learning activities. Each span should be the largest gain toward mastery that learners can make successfully. You want to come up with the smallest number of these spans that you can. Think in terms of “elegance” rather than “complexity” or “dumbing down.” At this point, I like to show my groupings to expert performers and to clueless beginners, as well as to other designers. With the right questions to each group to elicit feedback, I can find out pretty quickly whether my groupings are the appropriate size.
Sequence the instruction. The basic principle is to make sure that learners reach the goal or produce the desired outcome early. They should produce the desired outcome in every exercise or at least in every module of an e-Learning application. Over the whole application, the span of the learner’s activity will increase with each exercise until it includes the entire task. If in doubt, refer to figure 3 again.
Construct the exercises. Finally, you are ready to design the actual exercises. You may choose to be very thorough, as in Figure 3, or you may find it sufficient to show the learner what to do and then release them to do it on their own.
An exercise is a sequence of steps, each one of which provides a graduated level of help to move the learner toward mastery of the task. There are three levels of help that provide this support: demonstrate (sometimes called priming), prompt, and release (sometimes called performing).
In the context of Figure 3 and exercise construction, “demonstrate” can simply mean that the e-Learning application shows the learner what to do, through instructions or examples. It can also mean that the e-Learning application tells the learner what to do one little step at a time and the learner does each step immediately afterward. In either case the learner receives instructional aid from the e-Learning system to guide performance to a successful conclusion.
“Prompt” means that the e-Learning system provides hints or helps as needed by the learner. These helps can be in the form of partial hints or completely worked out examples. The system can do part of the work for the learner by providing partly worked out exercises, multiple choice questions, or pictures. Or the helps can be “memory joggers,” such as mnemonics or examples. This step can often be skipped, especially if it seems to be dumbing down the instruction or prolonging it unnecessarily. You have to know your audience and judge accordingly.
“Release” means that the learner does the task without any help from the e-Learning application. This is the step where transfer to long-term memory and to the job begins.
You will need to sequence the exercises using these levels. Figure 4 provides a general plan for teaching a task element that has three steps. Remember, if the plan in Figure 4 results in a sequence that is too tedious or boring for the learners, or one that insults the learners’ intelligence reduce and simplify to obtain a better experience for the learners.
FIGURE 4 Example of sequencing exercises for a three-step task element
What happens next
After you have developed the sequence and the exercises, you are in a better position to identify the overview and theory items learners will need. These items support the learner during the instruction and also support more generalized performance of the task back on the job.
At this point, you are ready to begin development of the e-Learning itself. If your organization uses storyboards, you have a very good basis for creating them. For simple e-Learning applications, the sequence and exercise designs may be sufficient if the designer is also doing the development. In any case, I recommend trying out the e-Learning through alpha and beta testing and making necessary adjustments before release.
Summary
When I teach instructional designers how to use backward chaining, the content of this article is the introduction and overview. The rest of the instruction is delivered as a backward chain, beginning with construction of exercises to deliver instruction for a job that has been described, grouped, and sequenced. This procedure works whether the delivery of my instruction to the designers is done in a classroom, in one-on-one coaching, or as an e-Learning application.
In this article, I have chosen not to go that far. Mainly this is in the interest of space, but it is also because the range of content suited to backward chaining is so vast that no one example would be a helpful illustration for every Guild member. I am happy to field simple questions by email, however.
You will find that backward chaining is useful whether your philosophical orientation tends toward the behavioral school or toward the cognitive school. It’s mainly a matter of choosing a sequence that keeps the forgetting curve in check, while advancing the learning curve. I hope you enjoy using this powerful tool in your e-Learning designs!
