Uyarlanabilir Dönüt Sistemi Tasarımı İçin Kullanıcı Profillerinin Belirlenmesi

Abstract

This research aims to create feedback adaptation models and rules for e-assessment systems according to the characteristics of the learner and the learning task. Accordingly, e-assessment tasks that require different types of knowledge are given to learners, the types of feedback that the learners prefer when performing these assessment tasks are investigated. The patterns related to the classification of learner feedback preferences are revealed based on the learner characteristics and type of knowledge required by the assessment task. Moreover, inter-feedback interactions were examined with sequential analysis methods in the context of feedback seeking strategies. The research is designed according to descriptive and correlational research methods. Firstly, the examination of the experiences of the learners in the e-learning environment shows that e-assessment is an important component of e-learning. Secondly, when learner e-assessment/e-feedback interactions were examined, it was seen that the knowledge of correct response and elaborated feedback types were preferred primarily. Besides, three different decision tree models have been created based on learner characteristics. When the decision trees are examined, it is seen that the level of learner knowledge is the most important classifier variable. While students with low prior knowledge tend to receive the knowledge of correct response feedback, those with a high level of prior knowledge tend to receive elaborated feedback. Lastly, lag sequential analysis results show that the learner sequential feedback patterns differ according to their prior knowledge, task value, self-efficacy, cognitive style, and cognitive learning strategies.