Agents in Online Courses 1 Running Head: AGENTS IN ONLINE COURSES Exploring the Use of Pedagogical Agents in Online Human Service Courses

Human service educators have long struggled with how to best provide students with the communication skills required of entry-level human service professionals. Traditional methods of teaching in this domain have included role-plays and case studies. While teaching such skills is easier in a traditional face-to-face environment, this issue has become more complex due to the emphasis on delivery of instruction at a distance. One way to solve this dilemma may be through the use of pedagogical agents embedded in a web-based learning environment. This presentation will describe the development of a web-based learning environment designed to help online human services students with these skills. We address the early implementation and comparison of three web-based environments. In the interactive environment, learners play the role of a human services helper while the animated agent portrays a client. In the modeling environment, participants observe a client-helper interaction between two agents. The text only environment presents the students a helper-client script to read. Data was collected to evaluate the usability of the environments and skill acquisition. Agents in Online Courses 3 Introduction A primary goal in training human service professionals (HSPs) is to give them effective helping skills to work with clients on a daily basis. HSP services vary depending upon the employment setting and the clients. Despite this broad array of settings, HSPs share a primary purpose: to assist individuals and communities in functioning as effectively as possible (National Organization for Human Services [NOHS], 2004). Communication skills are so important for the HSP that the Council for Standards in Human Service Education [CSHSE] identifies effective communication strategies and skills as one of the 12 skill sets that entryand mid-level human services workers use daily in their jobs (n.d.). Human service educators face constant challenges to provide learning experiences that assist students in developing their skills for use with clients, co-workers, and supervisors. Human service programs offered through distance learning face even more difficulties as educators struggle with designing online and interactive video courses that provide opportunities for helping skills practice along with the evaluation of such skills. Techniques used in traditional classes such as role-plays, group work, and even videotapes are much more difficult to adapt for use in a distance learning setting. Add to this the challenge of large class sizes often found in distance learning, and teaching interpersonal communication skills at a distance becomes even more overwhelming. To remedy this issue, educators may want to consider other less traditional methods to provide opportunities for their distance-learning students to learn and practice helping skills appropriate for the field. This idea is not without precedent as systems to supplement similar skills such as crisis intervention are being developed and tested (Seabury, 2003). This paper describes the evaluation of a system designed to provide online training in essential skills. Three Agents in Online Courses 4 systems were tested, two employing pedagogical agents and one script of a helper-client interaction. Pedagogical Agents in Distance Learning Pedagogical agents are computerized animated characters used as tutors in computerbased instructional environments. Such characters allow designers to create environments where learners can interact with a conversational partner to obtain advice, feedback, or instruction. Consistent with early findings of human-computer interaction, agents allow for a realistic social interaction (Reeves & Nass, 1996). By using advanced technologies such as artificial intelligence, agents can be designed to support or facilitate learning by providing a teacherstudent interaction with the learner (Shaw, Johnson, & Ganeshan, 1999; Slater, 2000). Pedagogical agents have been used in many capacities and have been shown to be effective in teaching, largely due to their impression of a realistic social interaction (Atkinson, 2002; Lester, Converse, Kahler, et al., 1997; Moreno, Mayer, Spires & Lester, 2001). Pedagogical agents are primarily designed so that a single user engages in an interaction with a single agent. However, some research focuses on learning when users are exposed to two types of agent environments: interactive environments where the learner directly interacts with the agent and vicarious environments where learners observe agent-agent interactions (Craig, Driscoll & Gholson, 2004). Findings from the aforementioned study indicated that learning gains were greater when learners interacted directly with the pedagogical agent (attributed to the active learning occurring during the interactions) while observance of a collaborative tutoring session produced smaller learning gains. The results from this study merit further investigation in domains where learners may benefit by observing effective interactions as a model of Agents in Online Courses 5 appropriate behavior, such as in human services, consistent with findings in social learning theory (Bandura, 1977). Some of the current uses of pedagogical agents include their presence in multimedia instructional environments teaching middle school botany (Moreno, Mayer & Lester, 2000), instructional planning (Baylor, 2001), computer literacy (Gilbert, Wilson & Gupta, 2005; Graesser et al., 1999), conceptual physics (Graesser, VanLehn, Rose, Jordan, & Harter, 2001), foreign language and culture for military personnel (Johnson, 2004), role playing (Prendinger & Ishizuka, 2001) and applications such as interviewing, negotiation, and patient assessment (Hubal & Guinn, 2001; Hubal, Frank & Guinn, 2003). In all of these instances, evaluations have supported the use of agents as a viable replication of a human-human interaction and as an effective tool in online learning. It is conceivable to think that another possible application for pedagogical agent environments is in teaching communication skills through the simulated interviews or through the vicarious observation of an interview between helpers and their clients. It is also important to compare these agent-inhabited environments with the presentation of an effective scenario via hypertext to determine that resources necessary for pedagogical agent development are well used. Purpose of this Study The purpose of this study was to investigate the use of pedagogical agents to assist human services students at a distance in practicing helping skills. The rationale for this investigation is the increasing call to deliver human services instruction to a large number of students at a distance. By determining if pedagogical agent learning environments are an effective delivery mode for practicing these types of skills, the researchers hope to add to the tools available for distance education in human services. In addition, this study will examine Agents in Online Courses 6 whether the type of instruction provided by the pedagogical agent (interactive or modeling ) is differentially effective. This study answered the following specific research questions: • Are web-based pedagogical agent environments effective tools for the practice of communication skills for online human services students? • Are there differences in skill acquisition when human services students interact with a pedagogical agent in a case study learning activity compared to reading text-based case studies? • Are there differences in skill acquisition when human services students observe an interaction between two pedagogical agents (modeling) compared to when they interact with the pedagogical agent in a case study learning activity (interactive)? Study Design/Methodology