Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Personalizing explanations of AI-driven hints to users' cognitive abilities: an empirical evaluation

Vedant Bahel, Harshinee Sriram, Cristina Conati

TL;DR

This work investigates personalizing the explanations that an Intelligent Tutoring System generates to justify the hints it provides to students to foster their learning and provides valuable insights into effectively personalizing AI-driven explanations for cognitively demanding tasks such as learning.

Abstract

We investigate personalizing the explanations that an Intelligent Tutoring System generates to justify the hints it provides to students to foster their learning. The personalization targets students with low levels of two traits, Need for Cognition and Conscientiousness, and aims to enhance these students' engagement with the explanations, based on prior findings that these students do not naturally engage with the explanations but they would benefit from them if they do. To evaluate the effectiveness of the personalization, we conducted a user study where we found that our proposed personalization significantly increases our target users' interaction with the hint explanations, their understanding of the hints and their learning. Hence, this work provides valuable insights into effectively personalizing AI-driven explanations for cognitively demanding tasks such as learning.

Personalizing explanations of AI-driven hints to users' cognitive abilities: an empirical evaluation

TL;DR

This work investigates personalizing the explanations that an Intelligent Tutoring System generates to justify the hints it provides to students to foster their learning and provides valuable insights into effectively personalizing AI-driven explanations for cognitively demanding tasks such as learning.

Abstract

We investigate personalizing the explanations that an Intelligent Tutoring System generates to justify the hints it provides to students to foster their learning. The personalization targets students with low levels of two traits, Need for Cognition and Conscientiousness, and aims to enhance these students' engagement with the explanations, based on prior findings that these students do not naturally engage with the explanations but they would benefit from them if they do. To evaluate the effectiveness of the personalization, we conducted a user study where we found that our proposed personalization significantly increases our target users' interaction with the hint explanations, their understanding of the hints and their learning. Hence, this work provides valuable insights into effectively personalizing AI-driven explanations for cognitively demanding tasks such as learning.
Paper Structure (11 sections, 6 figures, 3 tables)

This paper contains 11 sections, 6 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. The ACSP applet and its original explanation interface
  4. Personalized explanation interface
  5. User study
  6. Results for objective measures
  7. Effects on interaction with explanations
  8. Effects on learning gains
  9. Results on subjective measures
  10. Discussion
  11. Conclusion and future work

Figures (6)

  • Figure 1: (a) An example of ACSP hint; (b) The first page of the explanation interface that is activated by clicking on “Why I am delivered this hint?” at the bottom of the hint; (c) A schematic view of the six incremental explanations available.
  • Figure 2: (a) First explanation page appearing upfront with a hint; (b) Confirmation box.
  • Figure 3: Distribution of fixation duration (in seconds) for all deliveries of the first explanation page to users in the experimental group. D'n' refers to the n$^{th}$ delivery of the page for a given user.
  • Figure 4: Effect of personalization on PLG.
  • Figure 5: Subjective ratings of hints. "*" indicates a statistical difference between the control and experimental groups.
  • ...and 1 more figures