Immersive interfaces for spoken dialogue systems reduce learner speaking anxiety, both compared with a desktop environment, and when talking face-to-face after exposure

Thought process

  • There is a difference between learning a language and acquiring a language (Krashen)
    • Learning is to be able to use the language via conscious thought
    • Acquiring is to be able to use the language without conscious thought
  • Speech is the basis of language
    • Speech is an example of living language, evolving, etc.
  • Practising SLA speech is the best way to improve SLA speech
  • (1) Verbal reproduction of second languages during acquistion has a relationship with anxiety
    • Anxiety is caused by the process of using a second language
    • Anxiety affects the ability to use a second language (physical effects)
    • Anxiety affects motivation
    • Anxiety affects learning outcomes
  • (1) VR practice has been shown to lower anxiety in real-world situations
  • (1) Can VR practice do the same for verbal SLA?
  • (2) Practicing speech is also difficult
    • Difficult to do alone
    • Difficult for self-learners to find third-party experts to provide corrections
    • Difficult for classroom study as teacher interaction time is limited
  • (2) We need an on-demand system that could tell us if we are speaking correct SL phrases
  • (2)But, speech is dynamic and unpredictable. Rote practice is not enough to improve SLA speaking *citation needed
  • (2) We construct responses dynamically (from Universal Grammar), so any system must be dynamic in its understanding
    • Therefore a dynamic environment is needed

Environment questions

  • How does VR interaction occur?
    • Informed by Game Design theory
    • Informed by SLA theory
  • What is practiced inside the environment?
    • Informed by SLA theory
  • Can Google or Amazon services provide a high quality of service for verbal understanding?
    • Informed by testing
  • How do we comprehend and process meaningful inputs from users to outputs from the system?
    • Informed by Game Design theory (decision trees, machine learning, fuzzy logic)

Testing questions

  • How do we measure anxiety?
    • Informed by CogSci theory
  • How do we measure SLA outcomes?
    • Informed by SLA theory
  • How do we control for implicit VR factors, like interaction novelty?
    • Informed by CogSci theory
  • How do we control for embodied interaction impact?
    • Informed by testing
  • How do we control for ludic impact/intrinsic motication?
    • Informed by CogSci + SLA theory
    • Informed by testing
  • Will we need to try different aspects of speech to see success levels?
    • Informed by SLA theory
    • Informed by testing

Ideal participants

  • Classroom or self-language learners
  • People who have used VR before
  • People with an existing knowledge of the target language's phonetics