Learning Through Prompted-Experiences in Online Interfaces


IRB > Data Gating > Data Analysis  > Results > Publications


More and more students use computer interfaces for learning, and online material is frequently presented in a variety of ways. We are interested in identifying ways to maximize online learning. Specifically, our research explores how unique online interfaces influence reading and learning.


Past research has demonstrated that students who spontaneously self-explain and those that are prompted to self-explain when reading a textbook learn with a greater understanding than those who do not (Chi, Deleeuw, Chiu, & LaVancher, 1994). Further research determined that when students are prompted to self-explain through a typed response within a computer interface, the same quality of self-explanation is not elicited; instead a higher incidence of paraphrasing is seen (Hausmann & Chi, 2002). Use of elaborate self-explanation tutorial dialog (Alevan, Popescu & Koedinger, 2001) or an "SE-Coach" (Conati, Larkin & VanLehn 1997, Conati & VanLehn 2000) suggest that students learn better when some interactive scaffolding is provided to elicit self-explanation based on the idea that students can be taught to self-explain (Bielaczye, Pirolli & Brown, 1995). Considerable research has been made that transitions the concept of self-explanation to that of explanation to a Teachable Agent (TA), or a computer generated "student" (Brophy, Biswas, Katzlberger, Bransford, & Schwartz 1999, Blair & Schwartz 2007). In the case of our research, the purpose is to compare different types of intervention to self-explanation explicitly for potential learning affect. In addition to a control condition, three experimental conditions will be used. In the first condition, students will be taught to paraphrase and will be prompted to paraphrase material. In the second condition, students will be taught to self-explain and will be prompted to self-explain the same material. In the third condition, students will be asked to read the material and teach a TA about what they have learned. Through this direct comparison of different modalities to elicit self-explanation, it is hoped that specific ways to optimize learning through computer interface will identified.