System and Learner Control in Adaptive Systems

Types of Adaptive/Adaptable CAI Learning Systems

CAI can be designed as an adaptive system through which learners are placed at an appropriate learning level after taking a placement test, for example. The system determines the learning level of the learner through analyzing the learner’s test results. The learner takes instruction and then another test. The system keeps analyzing the learner’s test results and places the learner at a higher or lower learning level based on learning performance. This example is used to illustrate system control in a simple adaptive learning system. The placement of the learner is determined by the system. Enabling system control are programmed modules (usually based on subject matter experts’ insights) and if/ then rules, such as a module for generating test items based on a learner’s current level, a module for checking learning deficiencies (e.g., comparing responses to deficiency criteria), and providing a recommendation for the next learning level for that particular learner. It is not always necessary for the learner to complete all the items on a test for the system to determine an appropriate level of placement. As long as certain criteria are met while the learner takes the test, the system is able to determine where the learner should be placed.

A variety of common CAI learning systems are designed as learner control systems in which learners can choose a level or next module. In a learner-control CAI, the system is designed with branches at different levels with a variety of subareas. Learners usually do not take the placement test at the beginning because the system is not in control so the data will not be analyzed or utilized. Learners can access instructional modules (small units of instruction) at different levels to determine where to start. They can also change the level based on their own learning experience.

...