This study focuses on collaborative learning involving a knowledge integration activity, whereby learner dyads explain each other’s expert knowledge. It was hypothesized that learn- ing gain can be determined by the degree to which learn- ers synchronize their gaze (gaze recurrence) and use overlap- ping language (information overlap) during their interaction. Thirty-four learners participated in a laboratory-based eye- tracking experiment, wherein learners’ gazes and oral dialogs were analyzed. Multiple regression analysis was conducted, wherein learning performance was regressed on the two inde- pendent variables. Then, a simulation was conducted to view how the model predicts performance based on the collabora- tive process. The results showed that both gaze recurrence and lexical overlap significantly predicted learning performance in the current task. Furthermore, the suggested model success- fully predicted learning performance in the simulation. These results indicate that the two variables might be useful for de- veloping detection modules that enable a better understanding of learner-learner collaborative learning.

Collaborative problem solving based on different perspectivesis an effective strategy for constructing new knowledge anddiscoveries. It remains unclear what kind of interaction pro-cess underlies development of an abstract or integrated per-spective upon experiencing conflict with different perspectivesin a group. The present study investigates two factors in anexperimental setting: (1) groups with a single opposing per-spective (maverick) would hold an advantage over groups and(2) groups with positive moods would hold an advantage overgroups with negativity. We investigate the factors influencingperspective taking in problem-solving groups using conversa-tional agents. Results showed that (1) a single different per-spective in the group can be accepted for perspective takingcompared to several members with an opposing perspective,and (2) positive mood generated by group members facilitat-ing perspective taking compared to negative mood.

This study experimentally investigated the effect of sugges-tions from a physically present robot on human creative gener-ation. In the experiments, we used a creative task in which theparticipants were required to draw creatures living on a planetother than the Earth, and a physically present robot, which pro-vided suggestions for creative drawing to the participants withspeech sounds and physical movements. First, the results ofthe pilot experiment confirmed that drawing creativity was en-hanced for the participants supported by a robot; however, theywere unlikely to refer to the suggestions. Based on the re-sults, two hypotheses were developed: the suggestions from arobot offered a variety of different perspectives and facilitatedmetacognition (Hypothesis 1), and the suggestions worked asdistractions and suppressed fixated perspectives (Hypothesis2). The experiment was conducted to investigate these hy-potheses. As a result, Hypothesis 1 was supported. The resultswere discussed based on previous studies.

This study investigated how the expectations of others (i.e.,top-down processes) and actual perceived behavior (i.e.,bottom-up processes) influence negotiations during human-agent interactions. Participants took part in several sessions ofthe ultimatum game; we investigated the bargaining strategiesdirected toward the computer agent. To investigate the influ-ence of top-down and bottom-up processes on performance,we designed an experiment wherein (1) participants expectedtheir partners were humans or agents, and (2) agents used dif-ferent types of algorithmic behavior. Results revealed that ir-rational decisions, which are characteristic of human-humaninteractions, emerged when participants believed their oppo-nents were human and when opponent behaviors were ambigu-ous. Further, we found participants adopted different bargain-ing strategies according to their expectations and the agent’sspecific algorithmic behavior. We discuss interplay of the twotypes of cognitive processing in human-agent interaction.

This study investigated how participant’s specificity in shar-ing of information in collaborative problem solving was criti-cal to them reaching a successful shared perspective. We ana-lyzed participants’ communication strategies in a collaborativetask designed to make finding common ground challenging.We set out to better understand the difference between suc-cessful and unsuccessful collaborations by conducting a cog-nitive task analysis. From participants’ utterances, we inferredcognitive processes associated with repeating communicationmoves and coded those processes as if-then production rules.We thereby specified the communication strategies used duringinteractions and developed a production-rule model to explainwhether and how shared perspective developed or not. Ourcognitive task analysis indicated that although all collaboratingpairs described the objects they were seeing with a variety offeatures, the successful pairs were more specific in using com-binations of features. Quantitatively, we found significant cor-relations between frequency of combined feature statementsand success in sharing perspectives.

This exploratory study investigated how the combination of top-down and bottom-up processing influences decision-making for high and low trusters using the ultimatum game against a computer agent. We designed an experiment wherein (1) participants expected their partners to be humans or agents (top-down processes) and (2) agents used one of four different types of algorithmic behavior (bottom-up processes) to propose and respond. We found that high trusters made fairer decisions in the human condition than in the agent condition in the proposal phase, when opponents’ behaviors were not ambiguous but intentional. In the response phase, the higher the level of trust, the more likely they were to avoid unfairness for an opponent that proposed a distribution amount approved by the participant. Results suggest that, in interpersonal communication, high trusters flexibly use both types of cognitive processing to economically process information when developing representations of others and deciding on a response.