Academic journal article Educational Technology & Society

The Effects of Peer-Like and Expert-Like Pedagogical Agents on Learners' Agent Perceptions, Task-Related Attitudes, and Learning Achievement

Academic journal article Educational Technology & Society

The Effects of Peer-Like and Expert-Like Pedagogical Agents on Learners' Agent Perceptions, Task-Related Attitudes, and Learning Achievement

Article excerpt

Introduction

Pedagogical agents are virtual characters embedded in multimedia learning environment that simulate human instructional roles (Johnson et al., 2000). In today's multimedia technology, virtual agents are incorporated in several e-learning systems such as AutoTutor (McCauley et al., 1998), Herman the Bug (Barlow et al., 1997), and Steve (Rickel & Johnson, 2000). Advocates of pedagogical agent cited the benefits of the persona effect (Barlow et al., 1997), which posit that the social cues exhibited by pedagogical agents can increase learner's motivation (Barlow et al., 1997), cognitive engagement (Johnson et al., 2000; Mayer, Sobko, & Mayer, 2003), self-efficacy (Atkinson, 2002) and transfer achievement (Moreno et al., 2001) in learning tasks.

The persona effect is somewhat related to the idea of media equals real life in human-computer interaction. That is, people will tend to apply the same social rules of human to human communication to computer agents (Reeves & Nass, 1996). For example, it was shown that college students' perceptions on virtual agents were "knowledgeable, nice and friendly," which reflected the learners' social expectations of human instructors (Kim, 2007). Furthermore, users tend to perceive computer agents to be more appealing and intelligent when computer agents matched their personalities (Nass et al., 1995).

The activation of social expectations is triggered by the visual and voice characteristics of a virtual agent (Haake & Gulz, 2008; Kim et al., 2003; Veletsianos, 2010). Specifically, agent features such as voice inflection, hairstyle, clothing, ethnicity, and gender form the basis of first impressions on agent usefulness, credibility, and intelligence. A number of researchers have highlighted the importance of agent design (Moreno et al, 2001; Haake & Gulz, 2008). In a review done by Heidig & Clarebout (2011), it was concluded that the choice of agent design reflects the social perceptions such as competency and appeal, which in turn promote or hinder learning.

However, it seems that the importance of agent design has not been acknowledged accordingly in a number of studies. Veletsianos (2010) argued that researchers are failing to regard the effects of agent design on learner's stereotypic beliefs, which in turn, affect learning performance and task-related attitudes. In his study, the researcher cited some examples in which virtual agents were designed without considering the impact of visual stereotypes. For example, sorcerers have been employed to teach economics (Craig et al., 2002) and cartoon-like characters have been depicted as physics experts (Mayer, Dow, & Mayer, 2003).

One possible reason as to why agent design is often neglected in research is that much of the research on virtual agents is technologically driven; hence, the cognitive and psychology aspects are often disregarded. In addition, the choice of agent design is not an easy task (Haake & Gulz, 2008). The researchers cited an example of a virtual character that was designed for a legal information system in Italy (de Rosis et al., 2004). The virtual agent was originally modeled upon a very attractive young female assistant, since the developers anticipated that the typical user would be a male lawyer. However, after realizing that the lawyer's (female) secretary was the one who most frequently used the system, they became aware that the appearance and behavior of the character distracted these users. Therefore, the designers remodeled a new virtual agent with more professional communication style and more classical attire. From the cited example, it was apparent the choice of agent design is contextual to the subject domain and the target user.

Stereotypes of expert-like and peer-like virtual agents in multimedia learning

The objective of this research is to investigate the effects of expert-like and peer-like agent stereotypes on learning outcome and behaviors. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.