- Develop case study simulation scenarios & background materials
- Initial V.M. Server set up
- Install A.L.I.C.E. software
- Training in Superbot Software & A.I.M.L. modification procedures
- Creation and Development of the Virtual Role-play: Principal Investigator Dr. Jennifer Frank and Co-Principal Investigator Dr. Deborah Schussler, along with research assistants Emily Chukusky and Alex Collopy, will create and oversee the development of the artificial intelligence pedagogical agent to be used for training. This involves creating the actual content for the Artificial Intelligence Markup Language (AIML), personally testing the pedagogical agent, and using other methods of research to expand the AIML knowledge base. (Need to tie TK/TLT into this description as well)
- Formation of the Project Website: Acquiring a domain name, and creating content for landing page and other areas of the project website will be completed by members of the research team.
- Program Functionality: Authentication, server logging, transcript saving and exporting functions will be completed by T.K. Lee and the TLT Studio, with input from the research team. Completion of the first edition of the Superbot software will be made accessible on the project website, and internal testing by all members of the research team will be completed. Any major issues with the pedagogical agent will be modified before pilot testing.
- Pilot Testing: The research team will recruit 20 individuals to beta test the virtual role-play. Participants will complete a Paradigm for Dialogue System Evaluation (PARADISE) study, and their virtual role-play transcripts will be saved for analysis. Dr. Frank and Dr. Schussler will determine whether the virtual role-play meets acceptable performance standards. Transcripts will be coded using the coding scheme developed by Robinson, Roque, and Traum (2010) and problematic domains in agent responsiveness will be identified.
- Technology Evaluation and Revision: All members of the research team, along with input from the TLT Studio, will revise and modify the AIML and Superbot software based on findings from the Pilot Test results analysis. Revisions will be made before the implementation of the intervention.
- Dissemination of Pilot Findings: Results of pilot testing will be posted via a dedicated project website.
- Intervention: Teacher candidates (N=20) in CI405 Methods of Classroom Management will participate in a unit on effective communication and bully intervention. They will then practice their skills using virtual role-play. A control group will also participate in a unit on effective communication and bully intervention. They will then practice their skills using traditional classroom pedagogy (e.g., role play). All teacher candidates will videotape a live role-play performance. All teacher candidates will be taught by co-Principal Investigator Dr. Schussler.
- Analysis of Data: Research Assistants blind to condition, under the supervision of Principal Investigator Dr. Frank, will code the video to determine the efficiency and quality of conversations. Drs. Frank and Schussler will test for group differences between intervention and control groups on measures of self-efficacy in responding to classroom bullying, and quality of performance during live role play.
- Manuscript Preparation & Grant Preparation: All research team members will work to prepare manuscript and grant materials.
- Dissemination of Experimental Results: For the public, results from this project will be disseminated via a dedicated project website, highlights in prevention science spotlight newsletters, articles prepared for high-circulation professional magazines and newspapers, and participation in prevention research center speaker series, which is publicized throughout the university.