I am a cognitive scientist with primary interests in learning environments, human-computer interaction, and intelligent dialogue in tutoring systems. As a graduate student at the University of Memphis, I have been working in these areas of research on projects affiliated with the Institute for Intelligent Systems (IIS) and the Cognitive Science Laboratory. These projects have included interdisciplinary collaborative grants involving virtual tutoring environments, requiring expertise in psychology, computer science, learning sciences, computational linguistics, and education. My research focus is on learning/educational technologies, with branches extending into human-computer interaction (perceptual, cognitive, and social aspects), pedagogical strategies, tutorial dialogue/interaction, and mechanisms of feedback. My broader research strategy is to employ basic scientific findings from these branches to motivate a coherent and effective framework for any learning environment.
My future endeavors will be influenced by a number of pedagogical theories I have successfully used within my own research. Included among such diverse approaches are modeling-scaffolding-fading, inquiry learning, reciprocal training, case-based reasoning, adaptive feedback, and misconception remediation. These specific views on pedagogy have informed ongoing development of learning technology. I am interested in applying, integrating, and evaluating these sophisticated strategies within new technologies.
Learning Environments
The majority of my graduate research and training has focused on emerging learning technologies that are grounded in the cognitive and learning sciences. I have had a significant impact on the development, management, and evaluation of many projects: intelligent learning environments (Human Use Regulatory Affairs Advisor, interactive Strategy Trainer for Active Reading and Thinking, AutoTutor, Source Evidence Explanation and Knowledge Tutor, Why/AutoTutor), automated tools for analyzing natural language and discourse (Coh-Metrix, Question Understanding Aid, and AutoTutor Script Authoring Tool), and theoretical research involving various technologies (computer-monitored experiments, eyetracking studies). I have worked on intelligent learning systems in collaborations with other institutions (Carnegie Mellon University, University of Pittsburgh, MIT, University of Illinois at Chicago, Northern Illinois University, Chi Systems, Knowledge Analysis Technologies, Thoughtware Technologies, and the Department of Defense). The research foundation for most of these technologies has been based on an interaction principle, namely that more interaction (exchange of ideas through turn-taking activities) between the learner and system will increase deeper learning. These interactions have consisted of natural language dialogue, interactive simulation in virtual worlds, and more conventional channels of human-computer interaction. Our highly interactive systems create complex and dynamic environments that react adaptively to individual learners. Working with these various projects has provided me with different research experiences: knowledge engineer, interface designer, materials developer, cognitive task analysis, outcome evaluation, usability testing, project manager/supervisor, and research director.
Research at the Institute for Intelligent Systems attempts to tackle some of the “tough” problems common to learning environments. We have investigated specific patterns of users’ learning behavior within interactive simulations (Jackson, Olney, Graesser, & Kim, 2006), assessed the potential for latent semantic analysis to guide adaptive tutorial dialogue (Jackson, & Graesser, 2006; Jackson, Person, & Graesser, 2004), examined the role of affective states during learning (Graesser, Jackson, & McDaniel, in press), evaluated the effectiveness and perceptions of animated conversational agents (Graesser, Jackson, Ventura, Mueller, Hu, & Person, 2003; Moreno, Person, Adcock, Van Eck, Jackson, Marineau, 2002), and discovered limits of popular learning theories (VanLehn, Graesser, Jackson, Jordan, Olney, & Rosè, in press)
From these experimental findings, I feel that learning technologies will undoubtedly play an influential role in the future of education. It is up to researchers and designers to ensure a quality of teaching, training, and tutoring that will benefit various facets of society. These learning technologies are not seen as replacements, but more as enhancements to classroom instruction. My primary interests for future work revolve around investigating how theories in the cognitive and learning sciences might productively influence the design of emerging learning environments. This may include both building brand-new technologies, while evaluating and enhancing existing ones.
Human-Computer Interaction (HCI)
My research focus has been on the educational and pedagogical issues relating to human-computer interaction. We know that different interfaces afford different types of interaction, but what type of interaction is best suited for specific goals? My previous research has targeted the learning technologies created at the Institute for Intelligent Systems. I have played a major role in evaluating these learning environments for usability, functionality, and performance (Jackson, Olney, Graesser, & Kim, 2006; Graesser, Jackson, Ventura, Mueller, Hu, & Person, 2003; Jackson, Mueller, Person, & Graesser, 2001). My contributions have focused on learning gains resulting from interactions with the educational/ learning environments, on specific components of the interactions (tutorial structure, embedded tutorial simulations, dialog patterns of interaction, areas of attention focus within a system), on animated agents (perception, functionality, and value-add), on interface and tutorial feedback, and on motivation/engagement. My current research investigates different types of feedback (local and global levels of both progress and content feedback) and how they affect learning, motivation, and engagement within an intelligent tutoring system. This research incorporates aspects of cognitive psychology, education, pedagogy, and human-computer interaction.
Intelligent Dialogue in Tutoring Systems
Our research with intelligent dialogue systems is based on years of previous work which focused on human dialogue during one-to-one tutoring sessions. This previous research lead to theories of dialogue structure which have since been implemented within an intelligent tutoring system, called AutoTutor. Experimental findings using AutoTutor provided opportunities to examine some of the fine-grained aspects of dialogue sometimes hard to manipulate using human tutors. We have examined variety of specific combinations and sequences of tutoring moves (pumps, hints, prompts, assertions, discourse markers, short feedback, and question asking) and how these tutoring moves are related to student learning.
Research on AutoTutor has provided fertile ground for numerous research findings. Among them, we have worked with the University of Pittsburgh to conduct a large series of experiments on the learning of Newtonian physics. This research has shown that the interactive benefits of dialogue are more prominent when students have lower prior knowledge, and that the content of the material matters more than its method of presentation (VanLehn, Graesser, Jackson, Jordan, Olney, Rosé, in press). Additional research on AutoTutor dialogue has shown that it can accurately and adaptively respond to different student knowledge levels in pedagogically effective and conversationally appropriate ways (Jackson, Mueller, Person, Graesser, 2001; Jackson, Mathews, Lin, Graesser, 2003; Jackson, Person, Graesser, 2004; Jackson & Graesser, 2006). My work with AutoTutor, along with other technology from the IIS, has been used to investigate tutorial strategies (e.g., active knowledge construction, inquiry learning, case-based reasoning, etc.) and to compare theories of knowledge acquisition (Graesser, Jackson, Ventura, Mueller, Hu, Person, 2003; Jackson, Ventura, Chewle, Graesser, & TRG, 2004 ; Graesser, Lu, Jackson, Mitchell, Ventura, Olney, Louwerse, 2004; Jackson, Olney, Graesser, Kim, 2006; Vanlehn et al., in press). New versions of AutoTutor (including interactive simulations of conceptual physics scenarios) have shown that specific principle-relevant manipulations and increased exposure to simulations relate strongly with overall learning (Jackson, Olney, Graesser, Kim, 2006; Kim, Jackson, Graesser, & Sadkin, 2006).
I have worked on a wide range of projects that include systems which converse in natural language, provide hints and guidance through difficult content, have animated agents, that help students improve metacognitive/ metacomprehension strategies, and that analyze large volumes of text. My previous research experience has provided me with a broad and solid background within the cognitive and learning sciences. This allows me to pursue multiple research trajectories and to contribute to interdisciplinary research communities.