Using Low Interactive Animated Pedagogical Agents in Online Learning: An Exploratory Study of Singaporean Pre-Service Teacher Preparation

摘要

The aim of the study is firstly, to investigate the effectiveness of low interactive Animated Pedagogical Agents (APAs) in aiding learning for a sample of 378 Singaporean pre-service teachers taking a full-time Post-Graduate Diploma in Education course in a case Institute; and secondly, to gauge how favourably the pre-service teachers perceive learning through such instruction. The sample is chosen because it represents the largest cohort of students from the case Institute. The study also explores whether the effectiveness and favourability in regard to APAs are affected by learners’ sensory preferences. Because APAs are lifelike characters that can be embodied in a computer display to interact with learners, many APA-based lessons are designed with high interactivity to simulate intelligence; this type of APA however, is costly to develop and difficult to customise, making it less attractive for instruction. This study, in contrast, proposed that APA-based instruction should be designed with low interactivity, which is supported by sound pedagogies to help alleviate the above problem. To test this hypothesis, the study employed a quasi-experimental approach with a 2 x 4 factorial design to conduct the inquiry. Two learning conditions, the experimental and control conditions, and four sensory preference levels, the Strong Visual, Mild Visual, Strong Auditory and Mild Auditory levels - made up the factorial design. The two learning conditions were respectively learning with low interactive APAs (aka LIAI) and learning with a conventional online method (aka CI). Perceptions of LIAI were measured by three aspects of opinion: (1) extent of learning, (2) presentation of the instruction, and (3) interest in the instruction. At an overall level, 50% of learners on average were very positive in all three aspects of opinion; about 10% were very negative and the remaining 40% were mildly opinionated. At an individual level, extent of learning received the greatest satisfaction, followed by interest in the instruction and lastly presentation of the instruction. For effectiveness, LIAI produced moderately better learning performance than CI. Strong auditory and visual learners were found to learn best in their preferred modality. Mild visual learners learned well in their opposite modality and mild auditory learners did not benefit from either modality. The study also discussed implications of these findings and provided some recommendations for future research.