Abstract:

One of significant problems in self-directed learning is the lack of feedback from the authorities. Online questions and tasks, accompanied by the correct answers and explanations are good problem decision. But conventional types of test questions are countable and with high degree of repeatability, which follows to low results. In engineering learning area most of task types are based on calculations with strongly defined principles and algorithms. The input task data belongs to potentially infinite set of values and therefore the implementations tasks are potentially infinite set.

One task can be defined by set of input data, calculation logic and feedback text. Every input data can be defined via mandatory parameters: start and end value, and optional parameters: step and unit of measurement. Calculation logic can be described by: one expression; linear series of expressions; nonlinear series of expressions, containing branches or cycles. Describing the calculation logic requires some formal mechanisms, which must be suitable for person without programming skills.

Tasks are generated on run time. The real input values are random generated within definition area. Middle and output parameters are calculated according the described algorithm. User results are compared with calculated ones. Correct answers and corresponding expressions are shown as a feedback.

Self-directed assessment system is developing as a part of web portal for self-directed learning, experimentation and collaboration on projects in electronics and computer technology. Self-directed assessment system includes two sub-systems: task definition and task execution. One generated task type can be tested by author and after accurately labeled can be saved in the system for multiple executions.