Improving programming learners’ experience through interactive computer tutor based MOOCs

With the large demand for technology workers all around the world, more people are learning programming. Studies show that human tutoring is the most effective way to learn for novice programmers. However, problems such as the inaccessibility to physical classes, prohibitive costs, and the lack of educators may limit students\u27 opportunities to learn from these resources. Additionally, because programming is a skill requiring continuous practice and immediate feedback, simply listening to lectures may not be sufficient to learn effectively. This increases the inconvenience of learners who use online learning tools such as Massive Open Online Courses (MOOCs). In recent years, a particular type of MOOC that teaches programming in an interactive manner has become popular among programming learners, such as Codecademy and Treehouse. These systems are described as interactive computer tutors (ICTs) in this dissertation. ICTs provide an efficient solution for programming learners to practice the idea of learning-by-doing. The commercial application of ICTs has been growing rapidly in recent years and has gained a broad user base. Despite their success, there is limited research in the literature that addresses the users of ICTs. For example, who are the learners, what do they think of these ICT based MOOCs, and how can we improve their learning experience? This dissertation examines how learners interact with ICT based MOOCs and what design features improve their experience. Four studies were conducted to answer these questions. The results from these studies indicate that learners\u27 level of autonomy require different needs from ICTs, which are not addressed by current ICT designs. In addition, an autonomous feature is tested on an interactive computer tutor that teaches programming, and the effects for different learners are examined