COMET Presentation:
Interactivity in Computer-Aided Learning (CAL)

by Patrick Parrish
Cooperative Program for Operational Meteorology Education and Training (COMET),
University Corporation for Atmospheric Research (UCAR)

This paper is based on a presentation delivered at the Computer-Aided Learning in Meteorology (CALMet) Conference, 1995, Toulouse, France. Revised April 1996

Computer-aided learning (CAL) offers tremendous opportunities for educators and trainers to make their instructional materials more interactive. However, it is important that they have a plan for doing so, including a consideration of the appropriate levels, formats, and modes of interaction. To best apply interactivity, they should also understand why it is important to learning. This paper takes a brief look at the place of interactivity in the learning process and how it can be incorporated into CAL.

Information vs. Knowledge

It is useful for educators and trainers to make a distinction between "information" and "knowledge." We might think of information as referring to the initially inert facts, concepts, and rules that we ingest into memory, while knowledge refers to digested information and to an ability to employ that information to take action. Knowledge is the ability to use the information we possess to make decisions and solve problems that confront us. Information is simply one of the raw materials of knowledge. For that raw material to become useful, it must be processed through practice.

For information to be most useable, it is best acquired actively, not passively. The learner must do more than simply read or listen and then memorize. He or she must have the opportunity to practice using new information, preferrably applying it in realistic problem-solving situations. Many would even argue that information should be acquired in the context of problem-solving situations, rather than through a structure that separates information acquisition and practice (Brown, et.al. 1989). Instructional interactivity, the process in which students practice applying new information by answering questions or solving problems and then receive feedback, is an essential aspect of any effective instruction.

Why is interactivity important to knowledge acquisition?

Cognitive psychologists currently reject the metaphor of memory as an elaborate file cabinet as too simplistic. There are too many ways that we recall information for conceptual heirarchy to be the only method of organization. If the file cabinet were a good metaphor, then simply memorizing something under the correct category (i.e., filing it in the proper file folder and drawer) would be enough to build knowledge. Expertise would simply mean having many filled and well-organized file cabinets.

However, expertise is more than understanding the logical structure of a domain of knowledge. It also requires the ability to apply one's knowledge in unfamiliar situations, and even knowing when to throw out textbook procedures in favor of more expedient or effective solutions.

The more commonly accepted model of human memory is the "schema," a multi-dimensional web of interconnected "nodes" of information (Norman, 1982). According to schema theory, information is not stored in only one place or under only one category, as in a file cabinet. Instead, it has multiple connections to many other kinds of information such that it can be accessed via any number of paths. An individual schema is a group of nodes related in some useful, logical, or individually relevant way. Any one of the nodes, however, may be a part of other very different schemas. This multi-dimensionality makes memory richer and information more broadly applicable. The meteorological concept of vorticity, for instance, is applied in understanding many diverse meteorological phenomena, at both small and large scales. In other words, this concept is a node existing in and linking many different meteorological schemas (not to mention schemas used in other physical sciences).

In this model, the strength of knowledge relies not simply on the number of nodes that exist, but more on the quality and number of interconnections between the nodes. For example, expert weather forecasters are experts because even in unfamiliar situations they can quickly apply the appropriate meteorological concepts and principles in order to make a diagnosis and forecast. They are not experts simply because they know that those concepts and principles exist. An expert forecaster who has gained experience working in a high-plains environment can move to the southeast U.S. and still forecast thunderstorms, even though the meteorological signals may be quite different between the two climates. Such a forecaster can take the new climatology into consideration and mentally infer the implications of a phenomenon.

Figure A, right, represents a richly-interconnected schema, while Figure B represents one that is purely hierarchical.

Interconnections are built by using information to solve problems in unfamiliar situations and by comparing new information to existing information. In the context of instruction, interactivity encourages the deeper cognitive processing that builds interconnections. Our goal as educators and trainers should be the development of knowledge in our students rather than the dissemination of information. This means our instruction should contain ample amounts of interactivity. While these advantages of interactivity in instruction are fairly obvious, it is often used inadequately. It is mainly the practical constraints of delivering formal education to large groups that makes it difficult to use sufficient interactivity. Also, it takes time to develop rich, appropriate interactions. While lectures, demonstrations, and textbooks are efficient at disseminating information to large groups, without accompanying interactivity they are not particularly effective methods of developing knowledge that students can transfer to real world activities. The difficulties of learning a foreign language are a prime example. While classroom instruction can bring about the semblance of knowing a language, it isn't until one has the opportunity to communicate with native speakers of that language in real-life situations that fluency can be achieved.

The Interactive Nature of CAL

Interactivity is the cornerstone of computer-aided learning. CAL's capability to incorporate rich interactions between the student and the material sets it aside as a unique medium. Without interactivity, it is simply an amalgam of other media (text, graphics, video, audio, etc.). By tapping into the interactive nature of the computer, educators finally have the means to create effective interactive instruction that can be disseminated to a large audience.

Computer technology is not required for interactive instruction. Even a primarily passive textbook can incorporate questioning strategies to encourage deeper cognitive processing. Engaging classroom lecturers, especially those who uses a variety of media and pose problems to students, can create possibly the most interactive learning environment of all. What the computer contributes is the ability to create interactions that will be consistent for all learners, and yet individualized for each learner, whatever the size of the audience. When well-designed and executed, CAL can create the kind of dialogue between the learner and material that only a personal tutor could surpass.

How is Interactivity Incorporated into CAL?

There are several factors to consider when creating interactivity for CAL modules. These include (1) the capabilities of the authoring tools and the skills of the developers who use them, (2) the time available for development, and (3) the nature of the instructional objective. Together, these considerations will drive the character of the interaction.

Most CAL interactions can be classified in three dimensions: level of cognitive processing required, format of interaction, and mode of interaction. Each of these dimensions is elaborated below, with examples listed from simple to more difficult to implement. While these dimensions have a taxonomic usefulness, they are not specifically intended as a way to judge the quality of an interaction. A well-designed interaction is one that is appropriate for the instructional objective and the nature of the content.

Level of cognitive processing in learning

There have been many taxonomies created to outline the levels of cognitive processing useful for learning. The following list, adapted from Marzano, et al. (1988), is a summary of the work of many other theorists:

• Remembering (recalling)
  Example: List the types of data available from GOES-8 satellite imagery.
• Organizing (classifying, sequencing)
  Examples: List in order the steps for issuing a flash flood warning. State the defining characteristics of a Supercell storm.
• Analyzing (comparing/contrasting; identifying components, main ideas, patterns, or errors)
  Examples: Using radar data, identify possible Supercell storms. Using model data, locate areas of strong vertical motion.
• Generating (predicting, inferring)
  Example: Make a precipitation forecast based on the data presented.
• Integrating (summarizing, creating rules or new applications)
  Example: State which data products you would examine for the following forecast situations and why you would choose them.
• Evaluating (identifying strengths and weaknesses, verifying)
  Example: The forecast for Denver called for severe hail. What factors make you agree or disagree with this forecast?

It is important that the level of the interaction match the level of the instructional objective, and also important to ensure that objectives that call for more complex tasks are met by performing interactions that engage deep processing. While asking students to recall a list of information presented earlier (Remembering level) creates a basic level of processing, asking them to use that information to solve a problem creates deeper processing, and probably more permanent and useful learning. However, interactions requiring lower levels of processing can be useful to achieve certain sub-objectives of the instruction, those that build to the more complex main objectives.

Format of interaction

"Format" refers to the convention adhered to by the structure of the interaction. We can refer to both "micro-level" formats, those that describe individual interactions; and "macro-level" formats, or those that describe how interactions are used over the course of an entire unit or section of instruction.

Micro-level

• objective questions: e.g., multiple choice, true/false, fill in the blank
  (constrained answers; easy to evaluate)
• open-ended questions: e.g., longer written answers, annotating visuals by drawing on them
  (unconstrained answers; difficult to evaluate)
• note making: e.g., built-in opportunities to make notes
  (non-answers; evaluation may not be necessary, but tracking for individual learners can be difficult)
• navigation: e.g., the choice of path and sequence of information (can require complex development if the optimal selection of information is one of the skills being taught)
  

Macro-level

• drill and practice: e.g., repeated short, objective practice items
  (constrained choices, easier to develop)
• tutorial: e.g., presentation/practice/feedback pattern used repeatedly
  (constrained choices, focussed goals; easier to design)
• games: (where goals are extrinsic to the material) e.g., crossword puzzles, Jeopardy-like question sections
  (constrained choices, focussed goals; more difficult to design)
• simulations: e.g., displaced real-time case studies (guided or unguided)
  (unconstrained choices, broader goals; may be very difficult to design)

Mode of interaction (user/computer dialogue)

By "mode," we mean the method or medium used to carry out the interaction.

User-to-computer

• click to select (buttons or "hot spots"): easy to develop
• text entry (computer-evaluated short answer, learner/instructor evaluated open-ended): usually easy to develop
• click and drag (moving objects on screen): difficult, requires advanced authoring systems
• drawing on screen (evaluated or compare only): difficult or simple depending on method of evaluation

Computer-to-user

Micro-level:

• text: easy to design and develop
• graphics: more difficult depending on complexity
• audio: more difficult to produce
• video: difficult and costly to produce
• animation: difficult to develop

Macro-level:

• tutoring (feedback to specific responses): usually easier to design and develop depending on format
• coaching, tracking (feedback to a trend of responses): difficult to design and develop

The choice of format and mode of interactivity is constrained by available resources and time for development, but the choice of format and mode do not necessarily constrain the level of cognitive processing that can be demanded by the interaction. For instance, while objective question formats and button clicking modes are easier to develop, interactions that employ them can require even the deepest levels of cognitive processing if well-designed. A multiple choice question can ask students to evaluate, as well as remember. On the other hand, complex modes of interaction can be used to engage low as well as high levels of processing. This doesn't mean that complex modes should only be used for complex objectives, however, since the motivational effects of the more difficult to develop modes and formats are of value. Simple variety, too, has a positive impact on holding attention. Still, designers of instruction will want to consider their mix of modes, formats, and levels of interactivity in order to make judicious use of their resources.

REFERENCES

Brown, J.S., A. Collins, & P. Duguid (1989) Situated cognition and the culture of learning. Educational Researcher, 18(1), 32-41.

Norman, Donald A. (1982) Learning and Memory. New York: W.H. Freeman and Company.

Marzano, R.J., R.S. Brandt, C.S. Hughes, B.F. Jones, B.Z. Presseisen, S.C. Rankin, C. Suhor, (1988) Dimensions of Thinking: A Framework for Curriculum and Instruction. Alexandria, VA: The Association for Supervision and Curriculum Development, 68-114.