Mayer and McLaren - Social Intelligence And Computer Tutors

From LearnLab
Revision as of 16:01, 31 March 2008 by Bmclaren (Talk | contribs) (Findings)

Jump to: navigation, search

Building Social Intelligence into Computer-Based Tutors

Richard Mayer and Bruce M. McLaren

Overview

PI: Richard Mayer, University of California, Santa Barbara

Co-PI: Bruce M. McLaren, Carnegie Mellon University, Pittsburgh

Others who have contributed 160 hours or more:

  • John laPlante, Carnegie Mellon University, programming and website design and deployment
  • Brett Leber, Carnegie Mellon University, programming and website design

We have redesigned the lessons and tests of McLaren et al's stoichiometry tutor for the purpose of this new study. We are also in the process of creating “voice” versions of each lesson in which the tutor speaks using a friendly human voice, providing the student with hints and error feedback. We have recruited two high school chemistry classes totalling approximately 60 students to participate online in February and March, 2008. We have also made arrangements to test approximately 80 college students at the University of California, Santa Barbara in April and May, 2008 and will do another study in the summer of 2008. We plan to analyze the results and write a paper based on the studies in the late summer of 2008.

Abstract

The goal of this project is to examine how student learning is affected by social cues in computer-based learning environments, such as the conversational style of online cognitive tutors. In particular, students will learn how to solve stoichiometry problems in the Chemistry LearnLab, using a cognitive tutor that provides hints and feedback in direct style or in polite style (McLaren, Lim, Yaron, & Koedinger, 2007). The stoichiometry tutor has been used for other PSLC studies, in particular those by McLaren et al that have investigated personalization, politeness, and worked examples.

Our study is based on Brown and Levinson’s (1987) theory of politeness, which specifies how people create polite requests; Reeves and Nass’ (1996, 2005) media equation theory, which specifies the conditions under which people accept computers as conversational partners; and Mayer’s (2005) personalization principle in which people work harder to learn when they feel they are in a conversation with a tutor. Our working hypothesis is that learners work harder to make sense of lessons when they work with polite rather than direct tutors, because learners are more likely to accept polite tutors as conversational partners (Mayer, 2005; Wang, Johnson, Mayer, Rizzo, Shaw, & Collins, 2008).

Glossary

Research Questions

Do polite feedback and hints within a computer tutor lead to more robust learning than direct feedback and hints?

Does polite, audio feedback and hints within a computer tutor lead to more robust learning than text feedback and hints (whether polite or direct)?

Hypothesis

We have two hypotheses, based on these research questions, with the second built on the first:

H1
Students will experience more robust learning when they work with polite rather than direct tutors, because learners are more likely to accept polite tutors as conversational partners
H2
Students will experience more robust learning when they work with polite tutors that provide audio feedback and hints rather than polite or direct tutors that provide no audio feedback, because learners are more likely to accept audio polite tutors as conversational partners

Background and Significance

The polite tutor uses politeness strategies developed by Brown and Levinson (1978) in which the goal is to save positive face--allowing the learner to feel appreciated and respected by the conversational partner--and to save negative face--allowing the learner to feel that his or her freedom of action is unimpeded by the other party in the conversation. After interacting with the stoichiometry tutor on solving a series of problems for several hours, learners will be given a transfer test based on the underlying principles--including an immediate test and a delayed test. We expect learners who had the polite tutor to perform substantially better on the transfer test than learners who had the direct tutor.

We will also experiment with Clark & Mayer's Modality Principle, in which audio narration replaces onscreen text.

Independent Variables

The independent variables we will experiment with in our studies are politeness (either direct or polite) and audio (hints & feedback in audio or text).

These variables will be crossed, leading to a 2x2 factorial design with the following conditions.

  • Polite-Audio condition: Students work with the stoichiometry tutor that provides polite statements that are spoken
  • Polite-Text condition: Students work with the stoichiometry tutor that provides polite statements that are in text only
  • Direct-Audio condition: Students work with the stoichiometry tutor that provides direct statements that are spoken
  • Direct-Text condition: Students work with the stoichiometry tutor that provides direct statements that are in text only

Dependent Variables

Our plan is to include the following robust learning dependent variables in our studies.

  • Normal post-test: Students will take an immediate post-test, right after completing work with the stoichiometry tutor
  • Transfer: Conceptual, transfer questions will be included in the post-tests
  • Long-term retention: Students will take a second post-test, including conceptual, transfer questions, 7 days after the initial post-test

Findings

As discussed above, we do not yet have findings. We have recruited two high school chemistry classes for February and March of 2008 to do a two condition study (direct vs. polite) and have made arrangements to do a lab study at the University of California, using the same materials, in April, 2008. We will do the full 2 x 2 study discussed above (i.e., direct/polite crossed with audio/text) with students at the University of California in the summer of 2008. We expect to have first findings and write a paper at the end of the summer of 2008.

Explanation

This study is part of the Coordinative Learning cluster. The study follows the Coordinative Learning hypothesis that two (or more) sources of instructional information can lead to improved robust learning. In particular, the study tests whether an intelligent tutoring system with polite hints and feedback (in both text and audio form) leads to more robust learning.

Connections to Other PSLC Studies

  • This study has a clear connection to the McLaren et al study , in that both studies explore the effect of personalized, polite hints and feedback. In fact, it was through McLaren's original studies, built on earlier work on e-Learning principles by Mayer, that Mayer and McLaren decided to join forces.

Annotated Bibliography

  • So far, no papers have been published on this project. The collaboration between Mayer and McLaren began in the mid-Fall of 2007 and first published results are expected in the fall or winter of 2008.

References

  • Brown, P., & Levinson, S. C. (1987). Politeness: Some universals in language usage. New York: Cambridge University Press.
  • Mayer, R. E. (2005). Principles of multimedia learning based on social cues: Personalization, voice, and image principles. In R. E. Mayer (Ed.), The Cambridge handbook of multimedia learning (pp. 201-212). New York: Cambridge University Press.
  • McLaren, B. M., Lim, S., Yaron, D., and Koedinger, K. R. (2007). Can a Polite Intelligent Tutoring System Lead to Improved Learning Outside of the Lab? In the Proceedings of the 13th International Conference on Artificial Intelligence in Education (AIED-07), pp 331-338. [pdf file]
  • Nass, C., & Brave, S. (2005). Wired for speech: How voice activates and advances the human-computer relationship. Cambridge, MA: MIT Press.
  • Reeves, B., and Nass, C. (1996). The media equation. New York: Cambridge University Press.
  • Wang, N., Johnson, W. L., Mayer, R. E., Rizzo, P., Shaw, E., & Collins, H. (2008). The politeness effect: Pedagogical agents and learning outcomes. International Journal of Human-Computer Studies, 66, 98-112.