Comparison of Errorless Learning and Response Cost to simple Trial and Error learning in an automated learning environment

Abstract

Feedback is an important part of the learning cycle. Computers make immediate (corrective) feedback possible which increases the number of learning opportunities. Relatively little research has been done on feedback in automated learning environments. We believe that it is important to know what type of feedback is best to use, the timing of the feedback and how much feedback should be given. In this project we will be looking into the different types of feedback and if there is any difference in performance and enjoyment when learning with them. This project compared three different methods of feedback: Errorless Learning, Response Cost and Trial and Error, in an automated learning environment based on Discrete Trial Training. 24 nine year old children from two classes were randomly distributed to the three different feedback conditions. A comparison of how fast the participants learned to recognise and discriminate between three similar bacteria showed that the Response Cost condition was significantly slower that the other two conditions. A comparison of how enjoyable each feedback condition was showed little to no difference. This can be an important finding since many educational computer games use Response Cost as a feedback method for teaching.