The Influence of Voice on Pedagogical Agent’s Persona and Recall Performance

Abstract

Despite the prevalence of computer-generated speech, few studies have investigated the direct relationship between an agent’s voice and students’ perception or recall performance. This study investigated the effects of voice (without visual information) on students’ perception ratings and recall performance. Our results indicated that in the absence of visual information, students greatly preferred the human voice. The recall performance, however, indicated that the synthesized voices led to better recall performance. Implications for pedagogical practices are described in the discussion section.

Voice’s Effect on Student Effort Ratings and Recall Performance While Under Cognitive Stress

“…This conclusion would be consistent with the Social Agency Theory, which suggests that social cues provided by the pedagogical agent aid the learning process (Atkinson et al, 2005). In audio-only learning situations, participants rated the classic and modern voice engines as being worse pedagogical agents than the human voice (Craig & Schroeder, 2018;Morris & Chen, 2020). These results provide some evidence for the Social Agency Theory.…”

“…Participants also had the best combined-recall performance when the information was presented by the human voice condition than in either synthesized voice condition. A previous study by Morris and Chen (2020) found that participants had the best recall performance in the classic synthesized voice condition. Morris and Chen (2020) hypothesized that participants in the classic synthesized condition would have better recall when participants' cognitive load was not stressed.…”

“…A previous study by Morris and Chen (2020) found that participants had the best recall performance in the classic synthesized voice condition. Morris and Chen (2020) hypothesized that participants in the classic synthesized condition would have better recall when participants' cognitive load was not stressed. Participants in the current study were tasked with listening to the same audio lectures as the Morris and Chen (2020) study, but participants in the current study were placed under cognitive stress.…”

“…These results call into question past research that found the classic voice engine to be inferior to the human voice as a teaching agent (Mayer, 2017). Other studies have found that the voice effect may still exist in audio-only learning situations, where a classic voice condition produced better recall performance than a human voice (Morris & Chen, 2020). A clear consensus has not yet been reached on the effects of modern voice engines in audioonly learning situations.…”

“…A recent study by Morris & Chen (2020) also compared the classic voice engine, the modern voice engine, and a human voice in an audio-only learning situation and found that a voice effect may still exist in these situations. In the Morris and Chen (2020) study, participants rated the human voice as the best pedagogical agent, but participants had the best recall performance in the classic voice condition.…”