Reconsidering the voice effect when learning from a virtual human

Scotty Craig, Noah L. Schroeder

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

The current paper investigates an essential design component of virtual humans, the voice they communicate with, by examining the impact of varied voice types. A standard voice effect has held that human voices should be paired with virtual humans. The current study revisits this effect. In a randomized trial, virtual humans used one of three voice types (classic and modern text-to-speech engines, as well as human voice) to present information to a sample of participants from an online population. The impact of each voice type on learning, cognitive load, and perceptions of the virtual human were examined. The study found that the modern voice engine produced significantly more learning on transfer outcomes, had greater training efficiency, and was rated at the same level as an agent with a human voice for facilitating learning and credibility while outperforming the older speech engine. These results call into question previous results using older voice engines and the claims of the voice effect.

Original languageEnglish (US)
Pages (from-to)193-205
Number of pages13
JournalComputers and Education
Volume114
DOIs
StatePublished - Nov 1 2017

Fingerprint

learning success
cognitive learning
credibility
learning
Engines
efficiency
present

Keywords

  • Multimedia learning
  • Pedagogical agent
  • Synthetic voices
  • Virtual humans
  • Voice effect

ASJC Scopus subject areas

  • Computer Science(all)
  • Education

Cite this

Reconsidering the voice effect when learning from a virtual human. / Craig, Scotty; Schroeder, Noah L.

In: Computers and Education, Vol. 114, 01.11.2017, p. 193-205.

Research output: Contribution to journalArticle

@article{e05686fdb3164d37bf361e542fbe13dd,
title = "Reconsidering the voice effect when learning from a virtual human",
abstract = "The current paper investigates an essential design component of virtual humans, the voice they communicate with, by examining the impact of varied voice types. A standard voice effect has held that human voices should be paired with virtual humans. The current study revisits this effect. In a randomized trial, virtual humans used one of three voice types (classic and modern text-to-speech engines, as well as human voice) to present information to a sample of participants from an online population. The impact of each voice type on learning, cognitive load, and perceptions of the virtual human were examined. The study found that the modern voice engine produced significantly more learning on transfer outcomes, had greater training efficiency, and was rated at the same level as an agent with a human voice for facilitating learning and credibility while outperforming the older speech engine. These results call into question previous results using older voice engines and the claims of the voice effect.",
keywords = "Multimedia learning, Pedagogical agent, Synthetic voices, Virtual humans, Voice effect",
author = "Scotty Craig and Schroeder, {Noah L.}",
year = "2017",
month = "11",
day = "1",
doi = "10.1016/j.compedu.2017.07.003",
language = "English (US)",
volume = "114",
pages = "193--205",
journal = "Computers and Education",
issn = "0360-1315",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Reconsidering the voice effect when learning from a virtual human

AU - Craig, Scotty

AU - Schroeder, Noah L.

PY - 2017/11/1

Y1 - 2017/11/1

N2 - The current paper investigates an essential design component of virtual humans, the voice they communicate with, by examining the impact of varied voice types. A standard voice effect has held that human voices should be paired with virtual humans. The current study revisits this effect. In a randomized trial, virtual humans used one of three voice types (classic and modern text-to-speech engines, as well as human voice) to present information to a sample of participants from an online population. The impact of each voice type on learning, cognitive load, and perceptions of the virtual human were examined. The study found that the modern voice engine produced significantly more learning on transfer outcomes, had greater training efficiency, and was rated at the same level as an agent with a human voice for facilitating learning and credibility while outperforming the older speech engine. These results call into question previous results using older voice engines and the claims of the voice effect.

AB - The current paper investigates an essential design component of virtual humans, the voice they communicate with, by examining the impact of varied voice types. A standard voice effect has held that human voices should be paired with virtual humans. The current study revisits this effect. In a randomized trial, virtual humans used one of three voice types (classic and modern text-to-speech engines, as well as human voice) to present information to a sample of participants from an online population. The impact of each voice type on learning, cognitive load, and perceptions of the virtual human were examined. The study found that the modern voice engine produced significantly more learning on transfer outcomes, had greater training efficiency, and was rated at the same level as an agent with a human voice for facilitating learning and credibility while outperforming the older speech engine. These results call into question previous results using older voice engines and the claims of the voice effect.

KW - Multimedia learning

KW - Pedagogical agent

KW - Synthetic voices

KW - Virtual humans

KW - Voice effect

UR - http://www.scopus.com/inward/record.url?scp=85024384458&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85024384458&partnerID=8YFLogxK

U2 - 10.1016/j.compedu.2017.07.003

DO - 10.1016/j.compedu.2017.07.003

M3 - Article

VL - 114

SP - 193

EP - 205

JO - Computers and Education

JF - Computers and Education

SN - 0360-1315

ER -