• Version
• Download 119
• File Size 321.59 KB
• File Count 1
• Create Date August 2, 2018
• Last Updated August 2, 2018

Combining computerized adaptive practice and monitoring: the possibilities of self-organizing adaptive learning tools

In the past few years there has been a rapid and large-scale introduction of digital technology in schools, which creates many new possibilities for educational assessment. We present a new methodology for combining web-based computerized adaptive practice and monitoring. This methodology is an extension of standard techniques of computer adaptive testing and is based on the Elo rating system (Elo, 1978). This self-organizing system enables simultaneously tracking the development of abilities and the calibration of item difficulties, as estimates of person abilities and item difficulties are updated with every answered item. In addition, both responses times and accuracy are used in the computation of ability and difficulty estimates. This methodology was first implemented in a web-based progress-monitoring system for math (Math Garden), which was originally developed at the University of Amsterdam to meet both educational and scientific aims. Within Math Garden children practice their math skills by playing games with items matched to their ability level. At the same time, their progress is automatically being monitored and presented to their teachers. In this paper we will give an overview of the underlying ideas and discuss the research that has been carried out with the Math Garden dataset.Key words: computer adaptive testing, high frequent monitoring, web based, educational assessment

Attached Files