How experts change their (viewing) behavior when modeling a task to a novice

Emhardt, S. (Speaker), Jarodzka, H. (Speaker), Ellen Kok (Speaker), Christian Drumm (Speaker), Brand - Gruwel, S. (Speaker), Tamara Van Gog (Speaker)

Activity: Talk or presentation typesOral presentationAcademic

Description

Instructional videos are gaining popularity, for instance on online learning platforms. Indeed, videos that display an expert (‘s eye movements) while demonstrating how to perform a complex problem-solving task such as code debugging, has the potential to foster novices’ learning. To study the cognitive and perceptual challenges novices usually face during code debugging, our first aim was to investigate how novices’ debugging behavior and eye-movement patterns differ from those of experts. Experts showed shorter fixations in the code area, showed fewer transitions between the panels of the programming environement per click on the run button, tested the code less often, and debugged the code more linearly compared to novices. These expertise-related differences in attention allocation and debugging behavior suggest that novices might benefit from attention guidance of experts. However, following authentic expert behavior might be challenging for novices. This is why expert models are typically instructed to behave didactically, yet it is not known how this affects experts’ behavior. Thus, the second aim of this study was to explore how experts change their eye-movement patterns and mouse click behavior when explaining their task solution didactically. In comparison to experts’ regular debugging behavior, didactically behaving experts showed longer fixation durations, shorter saccade amplitudes, ran the code less often, performed more transitions between code and output when running the code and debugged the code more linearly. Given that experts clearly change their (viewing) behavior in order to didactically guide a learner’s attention, an interesting question for future research would be to investigate which expert instruction is most suitable for video-based modeling example for novice students.
PeriodAug 2019
Event title18th Biennial EARLI Conference for Research in Learning and Instruction: Thinking Tomorrow's Education: Learning from the past, in the present and for the future
Event typeConference
Conference number18
LocationAachen, Germany, North Rhine-Westphalia
Degree of RecognitionInternational

Keywords

  • Experimental study
  • educational technology
  • instructional design
  • higher education
  • programming
  • eye tracking
  • expertise