Humans continuously assess one another’s situational context, modify their own affective state, and then respond based on these outcomes through empathetic expression. Virtual agents should be capable of similarly empathizing with users in interactive environments. A key challenge posed by empathetic reasoning in virtual agents is determining whether to respond with parallel or reactive empathy. Parallel empathy refers to mere replication of another’s affective state, whereas reactive empathy exhibits greater cognitive awareness and may lead to incongruent emotional responses (i.e., emotions different from the recipient’s and perhaps intended to alter negative affect). Because empathy is not yet sufficiently well understood, it is unclear as to which type of empathy is most effective and under what circumstances they should be applied. Devising empirically informed models of empathy from observations of “empathy in action ” may lead to virtual agents that can accurately respond in social situations. This paper proposes a unified inductive framework for modeling parallel and reactive empathy. First, in training sessions, a trainer guides a virtual agent through a series of problem-solving tasks in a learning environment and encounters empathetic characters. The proposed inductive architecture tracks situational data including actions, visited locations, intentions, and the trainer’s physiological responses to generate models of empathy. Empathy models are used to drive runtime situation-appropriate empathetic behaviors by selecting suitable parallel or reactive empathetic expressions. An empirical evaluation of the approach in an interactive learning environment suggests that the induced empathy models can accurately assess social contexts and generate appropriate empathetic responses for virtual agent control. Categories and Subject Descriptors

Rapid organizational change is increasing the pressure on employees to continually update their skills and adapt their behavior to new organizational realities. Goal orientation is a promising motivational construct that may explain why some individuals adapt to change better. Unfortunately, the current goal orientation literature is in a state of conceptual and methodological disarray. This presentation reviews the goal orientation literature and identifies numerous conceptual ambiguities, including definitional inconsistencies, dimensional inconsistencies, and inconsistencies in the conceptualization of stability. These conceptual ambiguities result in a confusing array of goal orientation measures and manipulations and ultimately an incoherent empirical database. A dynamic self-regulation model of goal orientation, termed motivated action theory, is presented to integrate the various conceptual perspectives and to provide guidelines for future goal orientation research.

By definition, collaborative e-learning activities imply that participants perform their work together with other individuals, i.e., other participants, tutors, and teachers. Although there is significant empirical evidence that the cognitive processes that are necessary for learning and knowledge construction occur in social interaction, and that “collaborative learning ” is the “royal road ” to knowledge acquisition, putting two or more people in the same context is not a warranty neither that they will be able to collaborate, nor that they will be able to learn. From the very beginning of research in social psychology (Triplett, 1897) it is well established that the mere presence of another is enough to modify the way an individual works, and a huge amount of empirical results have been collected: facilitation, competition, collaboration, group dynamics are only few examples of notions put forward for understanding the variety of phenomena documented; some of them if favor of beneficial effects, other against naïve interpretations that two people are better than one. As in many other phenomena both in everyday life and science, on-off interpretations are misleading if not wrong. At the same time of the beginning of 20 century, the cultural historical approach to development and learning was proposed. it is not room today for discussing the vicissitudes the how and why this

We have developed a software application (gStudy) that supports learning with multimedia documents. Students use gStudy to create and link notes, highlight and label text and images, construct glossaries and concept maps, exchange information objects through a chat interface, and perform other operations on multimedia content. The detailed log files recorded by gStudy constitute a wealth of data about how students process information as they learn. In this paper we describe log parsing and data mining methods we have borrowed from computer science and apply to researching selfregulated learning. The analysis software we developed identifies coherent learner actions from the complex series of low-level events recorded by gStudy and detects sequential patterns of these actions that may be interspersed with unrelated actions. It applies a data mining algorithm to discover action patterns which are the longest subsequences common to a group of participants. The use of these methods is illustrated through an analysis of gStudy log files generated by 103 university students. 2 Log Analysis of Student Interactions with Cognitive Tools

Affect has been the subject of increasing attention in cognitive accounts of learning. Many intelligent tutoring systems now seek to adapt pedagogy to student affective and motivational processes in an effort to increase the effectiveness of tutorial interaction and improve learning outcomes. To this end, recent work has begun to investigate the emotions experienced during learning in a variety of environments. In this paper we extend this line of research by investigating the affective transitions that occur throughout narrative-centered learning experiences. Further analysis differentiates the likelihood of affective transitions stemming from pedagogical agent empathetic responses to student affect.

Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author and do not necessarily reflect those of the National Science Foundation. Keywords Adaptive collaborative learning support, intelligent collaborative learning support, intelligent tutoring systems, reciprocal peer tutoring, computer-supported collaborative learning, collaboration scripts, in vivo Collaborative activities have been shown to be beneficial, provided that students exhibit certain positive behaviors. Unfortunately, these behaviors rarely occur spontaneously. Adaptive collaborative learning support (ACLS), where an intelligent system assesses student collaboration as it occurs and provides assistance when necessary, is a promising area of research that can help scaffold student collaboration. Little is known about how to build these adaptive systems and what effects they might have on collaboration and domain learning. In this dissertation research, I first augmented an existing intelligent tutoring system with a peer tutoring activity and then iteratively designed, built, and evaluated adaptive support for the activity. This dissertation focuses on two broad research questions: (1) Where and how can intelligent tutoring approaches be applied to the development of ACLS, and (2) Are there benefits to using existing

Abstract Recent studies have tested the addition of worked examples to tutored problem solving, a more effective instructional approach than the untutored problem solving used in prior worked example research. These studies involved Cognitive Tutors, software designed to support problem solving while minimizing extraneous cognitive load by providing prompts for problem sub-goals, step-based immediate feedback, and context-sensitive hints. Results across eight studies in three different domains indicate that adding examples to Cognitive Tutors is beneficial, particularly for decreasing the instructional time needed and perhaps also for achieving more robust learning outcomes. These studies bolster the practical importance of examples in learning, but are also of theoretical interest. By using a stronger control condition than previous studies, these studies provide a basis for refining Cognitive Load Theory explanations of the benefits of examples. Perhaps, in addition to other reasons, examples may help simply because they more quickly provide novices with information needed to induce generalized knowledge.