Return to Training Resource Center

A Brief Overview of the
Instructional Design Process

by Pat Parrish and Marianne Weingroff (COMET)

The Instructional Systems Design (ISD) process was developed to take the guesswork out of instruction. Rather than rely on assumptions or tradition about what needs to be taught and how best to present it, ISD takes a "systems" look at teaching. Among other things, it takes into account the environment in which learners are expected to eventually perform, characteristics of the learners, and aspects of the learning environment that could impact the effectiveness of the instruction. ISD is also a systematic process, one that suggests a series of steps that probe for the appropriate purpose and approach to the instruction and help guarantee effectiveness through intermediate evaluation and careful implementation. In these ways, the IDS method helps ensure that the instruction will actually be able to solve an identified problem or achieve a desired goal. Obviously, since training a large workforce is a costly effort, this kind of care in approach is important.

There are many models of the ISD process. Many theorists have suggested their own variations to emphasize various aspects of the process, but it is relatively easy to distill a generic version that gets at the main elements common to nearly all ISD models. One such generic model is the ADDIE process. ADDIE stands for Analysis, Design, Development, Implementation, and Evaluation, and about the only thing anyone would quibble with about this very basic outline is that the model shouldn't be construed as representing a strict chronological sequence of steps. For example, Evaluation might best happen at many stages in the process, and Analysis and Design could again come into play if Evaluation reveals problems in the process. As we outline each step in the ADDIE process you'll see many suggestions that are just common sense, but others that are not so obvious.

The ISD process is useful no matter what scale of instruction is being developed. Obviously, when the scale is small, as in a short presentation or a few pages of Web content, it doesn't have to be done as formally as in a project to develop several hours of or days of instruction. But the steps and prescriptions for what to consider are just as applicable.

Analysis

The Analysis phase includes several types of investigations that help determine:

1. Whether a perceived problem is related to lack of training or has a different cause. For example, a performance problem could be due to a problem in the work environment or in the job process itself. If so, training won't help.
2. Where performance gaps exist (i.e., job tasks forecasters could be expected to perform better), what job tasks require training, what the learners already know about the subject, and what misconceptions they might have, etc.
3. What type of training (instructor-led, self-paced, one-to-one) would be most appropriate for the environment. If, for instance, the training is to be done in a busy office with many interruptions, it may be best to design materials that are flexible and easy to pick up and put down. If forecasters are rarely available for group meetings, it may be best to create self-paced instruction followed by one-to-one discussions.
4. What the learners' preferred learning modes are
5. What physical facilities and equipment are available for training

A good analysis can be seen as fundamental to the whole ISD process. Without a careful determination of what tasks need to be trained and how they could be implemented, we can end up training the wrong things or things that will have less impact to the organization.

Numbers 1 and 2 above are the goals of what is known as the Needs Assessment. The PDS system that is rapidly ramping up in the NWS is in one sense a method for performing a systematic needs assessment. A PDS outlines the entire list of job tasks required for particular job responsibilities in order to create a useful starting point for taking the next steps--identifying the performance gaps and defining training priortities.

Questions:

• How do you currently decide when a problem needs to be addressed via training or some other means?
• What do you consider when deciding on a training approach?

Design

Once the training needs have been determined, it's time to move on to the Design phase, where the detailed decisions are made about what exactly will be taught and what approaches will be used.

First, an analysis is done to determine not just what information is required to perform the task, but also HOW the task is performed. The latter component (Task Analysis) involves direct observations and/or an attempt to "look inside the mind" of an expert in order to determine what thought process they go through and what factors they consider in making decisions. A formal task analysis may look like a flowchart. The former component (Content Analysis) focuses on the background knowledge needed to perform the tasks, considering the heirarchy of concepts, principles, and facts that add up to the prescribed body of knowledge. However, unless we're talking about a purely academic environment, we can't forget to consider the task processes as well. We always need to ask, "How will learners USE this knowledge." Otherwise, we could be presenting information in a context that is not useful or be unclear as to what depth to treat the information. Once the Task and Content Analysis is complete, further analysis must take place to determine what steps in the process or parts of the background knowledge are necessary to address. Obviously, not all learners are starting from scratch.

The next step is to develop the Objectives, which will be a crucial input for the rest of the process. Most commonly, these are written as Performance or Learning Objectives--those which describe what we expect learners to be able to do AFTER the instruction. For example, in the MCS Web module, after the analysis phase, we wrote the following objective: "Given existing convection, along with an updated hodograph and sounding, describe the potential for upscale organization and potential severity and longevity." This wasn't just an academic exercise. We referred to our objectives frequently to help drive the following steps in the process. For one, this objective (one among several) helped to narrow the scope of our instruction considerably. It reminded us that we did not intend to dwell on the initiation of convection, but only on those things which influence its upscale organization. This objective also helped determine what types of embedded and final exercises we were going to develop.

It may sound strange, but most ISD models suggest that the test design should take place after the objectives are defined, but before the actual instruction is created. Yes, it is suggested that you "teach to the test." If the objectives are well defined to begin with, this doesn't so much limit the amount of learning that takes place (as goes the valid argument about "teaching to the [standardized] test" in public schools) as ensure that the instruction stays on task and actually prepares the learner to be able to do what is intended. For the MCS module, we did not complete the final exercises before we created the rest of the module, but we did always have in mind what form they would take, and we gathered case data very early in the development process so we would have material to build the exercises. Many of the exercises took their form directly from the above objective, e.g., they provide soundings and hodographs along with radar data showing existing convection and ask leaners to predict how the convection will evolve. The final exercises perform at least two functions. They test whether learners have grasped the material and provide practice using the skills we hope they will gain.

The other steps in the design phase are to select an appropriate mode or medium (e.g., discussion, Web site, teletraining) and, of course, to determine what instructional strategies you will use. To some extent, these go hand-in-hand, as different modes are more or less conducive to certain strategies. For example, only highly interactive modes, such as computer-based instruction, classroom labs, or multimedia teletraining, are suitable for case-based instruction. Ideally, you would decide what instructional strategies you would like to implement and then choose an appropriate mode, not the reverse. But, of course, we are somewhat limited in our options due to practical constraints, so they do drive instructional strategy selection a bit.

Strategies can be both at a macro level and a micro level. Some possible macro-level strategies include tutorials (instructional presentations with frequent reinforcing exercises), direct presentation, discussions, games, or case-based instruction, where information is presented in the context of a goal- or process-based exercise. Some "micro-strategy" decisions include choosing to ask questions to draw knowledge out of the students vs. to directly present information, or deciding when and how to illustrate a concept with a visual. Strategy selection is a broad topic and theorists have many opinions about the best choices for particular instructional objectives. Perhaps these differences could be a topic of discussion for another thread.

Questions:

• Can you think of training you've seen or experienced in which the information or treatment was not aligned with what was required to perform the task?
• How do you decide on the "right" way to perform a task?
• When you are an instructor, do you have preferred instructional strategies to use?
• What strategies do you prefer as a student? Why?

Development

The Development phase refers to the stage of the project in which the materials are actually created. It includes such activities as preparing lecture notes, writing scripts, creating illustrations, coding Web pages, recording audio, etc.. Since so much of this phase depends on the medium, it is difficult to discuss it in general terms. However, two broad suggestions are to create and test prototypes early in the process and to have incremental reviews of the instructional materials from experts and people representative of the learners to insure that you are on track. The concept of incremental reviews (or Formative Evaluation) is extremely important within the ISD process. The basic premise is that instruction is one of those processes that is simply too complex for developers to know all the right answers in advance. (And you thought the weather was a "chaotic system"!) We can always learn how to improve the instruction when we do some test runs, even with small pieces as they are in development.

Questions:

• Do you "try out" your training before delivering it? Do you wish others did so?
• What new modes of training would you like to use? What support and assistance would you need?

Implementation

Successful implementation of instruction is not solely a function of well-designed materials, but requires having a plan for how they will be implemented and a structure for supporting the implementation. Although implementation plans differ depending on the type of instruction (e.g., small group, computer-based, etc.), they should all identify:

• When and where the instruction will be delivered and announced
• What equipment and facilities are needed
• What preparatory activities are required for the trainers and/or learners
• How learners will be evaluated or their participation monitored
• How field feedback will be gathered for potential revisions
• What follow-up activities will be provided to help insure that learning and transfer have taken place

The two final points are worth additional comment. It is highly desireable to have a mechanism for field feedback to clear up any confusion in the instruction and fix bugs in the implementation. The best mechanism is a Beta test with a small sampling. This takes time and delays the final delivery, but is well worth the trouble if it creates an improvement for the rest of the audience.

The goal of any instruction is to make a positive change in the learner's performance. Yet, because instruction exists more or less separate from the actual performance environment, we need a bridge to ensure that the knowledge and skills actually become incorporated into the regular job performance. This does not happen automatically, so special care must be taken during the design and implementation planning so that such a bridge exists. One of the key, but not only, ways that this should happen in the NWS is through the actions of the SOO. The SOO is necessary to provide the coaching and follow-on training opportunities to ensure that the instruction becomes ingrained in the learner. Since this is a large topic that generates much debate, we plan to hold another threaded discussion on the topic of designing and implementing for training transfer in the very near future.

Questions:

• Can you think of examples of training that suffered because of poor implementation? What were the causes?

Evaluation

The Evaluation referred to here is not a learner evaluation but more an evaluation of the instruction. This is often called a Summative Evaluation, since it happens after the instruction has been implemented. The goal is to evaluate the effectiveness of the instructional materials or program in order to make decisions about whether the instruction should be continued as is or designed and implemented differently. This can and should be done at several levels. You should not simply ask the learners and facilitors how they LIKE the instruction. This can get at some of the issues of design and implementation, but obvously not at whether the instruction actually does its job. Likewise, if you only ask "Did they learn from the instruction?" by presenting the learners questions drawn from the material, you have not gotten at whether the instruction actually was or can be transferred to the job. In the end, you also need to look at whether the job performance has changed, and whether this change has improved the quality of the work as a whole. Obviously, these are a lot more difficult to evaluate.

We propose that evaluation of both the learners and the training itself be another topic that deserves its own thread. If evaluation is to be done well in the NWS, it needs the support and involvement of the SOO.

Questions:

• Do you systematically evaluate the training you create? In what ways have you modified your training over time based on evaluation feedback?

Summary

The ISD approach to developing instruction provides a structure that helps ensure that you've considered all aspects of the instructional environment and maintained a focus on useful objectives. It need not be viewed as a lock-step, formal procedure for doing all instructional development projects. However, if you fail to address the intent of the steps in the process, you risk producing instruction that will not do what you intend it to do.