scispace - formally typeset
Search or ask a question
Proceedings ArticleDOI

"Oh dear stacy!": social interaction, elaboration, and learning with teachable agents

TL;DR: Treating her as a partner, primarily through aligning oneself with Stacy using pronouns like you or the authors rather than she or it significantly correlates with student learning, as do playful face-threatening comments such as teasing, while elaborate explanations of Stacy's behavior in the third-person and formal tutoring statements reduce learning gains.
Abstract: Understanding how children perceive and interact with teachable agents (systems where children learn through teaching a synthetic character embedded in an intelligent tutoring system) can provide insight into the effects of so-cial interaction on learning with intelligent tutoring systems. We describe results from a think-aloud study where children were instructed to narrate their experience teaching Stacy, an agent who can learn to solve linear equations with the student's help. We found treating her as a partner, primarily through aligning oneself with Stacy using pronouns like you or we rather than she or it significantly correlates with student learning, as do playful face-threatening comments such as teasing, while elaborate explanations of Stacy's behavior in the third-person and formal tutoring statements reduce learning gains. Additionally, we found that the agent's mistakes were a significant predictor for students shifting away from alignment with the agent.

Content maybe subject to copyright    Report

Citations
More filters
Proceedings ArticleDOI
08 Jun 2018
TL;DR: A study with 16 first-time chatbot users interacting with eight chatbots over multiple sessions on the Facebook Messenger platform revealed that users preferred chatbots that provided either a 'human-like' natural language conversation ability, or an engaging experience that exploited the benefits of the familiar turn-based messaging interface.
Abstract: Text messaging-based conversational agents (CAs), popularly called chatbots, received significant attention in the last two years. However, chatbots are still in their nascent stage: They have a low penetration rate as 84% of the Internet users have not used a chatbot yet. Hence, understanding the usage patterns of first-time users can potentially inform and guide the design of future chatbots. In this paper, we report the findings of a study with 16 first-time chatbot users interacting with eight chatbots over multiple sessions on the Facebook Messenger platform. Analysis of chat logs and user interviews revealed that users preferred chatbots that provided either a 'human-like' natural language conversation ability, or an engaging experience that exploited the benefits of the familiar turn-based messaging interface. We conclude with implications to evolve the design of chatbots, such as: clarify chatbot capabilities, sustain conversation context, handle dialog failures, and end conversations gracefully.

213 citations

01 Jan 2001
TL;DR: In this paper, the authors developed computer programs called PALs (Personal A_ssistants for L_earning) in which computers and students alternately coach each other.
Abstract: Our attempts to improve physics instruction have led us to analyze thought processes needed to apply scientific principles to problems—and to recognize that reliable performance requires the basic cognitive functions of deciding, implementing, and assessing. Using a reciprocal-teaching strategy to teach such thought processes explicitly, we have developed computer programs called PALs (P_ersonal A_ssistants for L_earning) in which computers and students alternately coach each other. These computer-implemented tutorials make it practically feasible to provide students with individual guidance and feedback ordinarily unavailable in most courses. We constructed PALs specifically designed to teach the application of Newton’s laws. In a comparative experimental study these computer tutorials were found to be nearly as effective as individual tutoring by expert teachers—and considerably more effective than the instruction provided in a well-taught physics class. Furthermore, almost all of the students using the PALs perceived them as very helpful to their learning. These results suggest that the proposed instructional approach could fruitfully be extended to improve instruction in various practically realistic contexts.

140 citations

Journal ArticleDOI
TL;DR: The overlap between HCI and sense of agency for computer input modalities and system feedback, computer assistance, and joint actions between humans and computers is explored.
Abstract: The sense of agency is the experience of controlling both one's body and the external environment. Although the sense of agency has been studied extensively, there is a paucity of studies in applied "real-life" situations. One applied domain that seems highly relevant is human-computer-interaction (HCI), as an increasing number of our everyday agentive interactions involve technology. Indeed, HCI has long recognized the feeling of control as a key factor in how people experience interactions with technology. The aim of this review is to summarize and examine the possible links between sense of agency and understanding control in HCI. We explore the overlap between HCI and sense of agency for computer input modalities and system feedback, computer assistance, and joint actions between humans and computers. An overarching consideration is how agency research can inform HCI and vice versa. Finally, we discuss the potential ethical implications of personal responsibility in an ever-increasing society of technology users and intelligent machine interfaces.

134 citations


Cites background from ""Oh dear stacy!": social interactio..."

  • ...Such agents have been investigated in application areas including education (Cassell, 2004; Ogan et al., 2012), healthcare (Bickmore and Gruber, 2010) and entertainment (Lim and Reeves, 2010)....

    [...]

Proceedings ArticleDOI
19 Apr 2018
TL;DR: By studying a field deployment of a Human Resource chatbot, data is reported on users' interest areas in conversational interactions to inform the development of CAs, and rich signals in Conversational interactions are highlighted for inferring user satisfaction with the instrumental usage and playful interactions with the agent.
Abstract: Many conversational agents (CAs) are developed to answer users' questions in a specialized domain. In everyday use of CAs, user experience may extend beyond satisfying information needs to the enjoyment of conversations with CAs, some of which represent playful interactions. By studying a field deployment of a Human Resource chatbot, we report on users' interest areas in conversational interactions to inform the development of CAs. Through the lens of statistical modeling, we also highlight rich signals in conversational interactions for inferring user satisfaction with the instrumental usage and playful interactions with the agent. These signals can be utilized to develop agents that adapt functionality and interaction styles. By contrasting these signals, we shed light on the varying functions of conversational interactions. We discuss design implications for CAs, and directions for developing adaptive agents based on users' conversational behaviors.

73 citations


Cites background from ""Oh dear stacy!": social interactio..."

  • ...Other studies showed that students engaged in playful interactions such as making face-threatening comments with tutoring agents, and found them to improve learning experience [32]....

    [...]

  • ...Recent studies considered this kind of behaviors as playful interactions and a key aspect of the adoption of CAs [28, 32, 43], through which users explore the system and seek satisfaction from a sense of social contact....

    [...]

Proceedings ArticleDOI
04 Jun 2016
TL;DR: A 17-day field study of a prototype of a personal AI agent that helps employees find work-related information is conducted and it is found that user differences in social-agent orientation and aversion to agent proactive interactions can be inferred from behavioral signals.
Abstract: Personal agent software is now in daily use in personal devices and in some organizational settings. While many advocate an agent sociality design paradigm that incorporates human-like features and social dialogues, it is unclear whether this is a good match for professionals who seek productivity instead of leisurely use. We conducted a 17-day field study of a prototype of a personal AI agent that helps employees find work-related information. Using log data, surveys, and interviews, we found individual differences in the preference for humanized social interactions (social-agent orientation), which led to different user needs and requirements for agent design. We also explored the effect of agent proactive interactions and found that they carried the risk of interruption, especially for users who were generally averse to interruptions at work. Further, we found that user differences in social-agent orientation and aversion to agent proactive interactions can be inferred from behavioral signals. Our results inform research into social agent design, proactive agent interaction, and personalization of AI agents.

66 citations

References
More filters
Journal ArticleDOI
TL;DR: This paper examined the relative impact of structured peer tutoring and group reward components of the reciprocal peer-tutoring intervention on the mathematics performance of elementary school students at high risk for academic failure.
Abstract: This study examined the relative impact of structured peer tutoring and group reward components of the reciprocal peer-tutoring intervention on the mathematics performance of elementary school students at high risk for academic failure. Sixty-four students were selected randomly from a pool of 80 4th- and 5th-grade students. Students were assigned randomly to four conditions: structure plus reward; reward only; structure only; and no structure, no reward. Findings indicated that students who received both components showed the highest levels of accurate mathematics computations

268 citations


""Oh dear stacy!": social interactio..." refers background in this paper

  • ...HYPOTHESES Cognitive hypotheses of learning by teaching suggest that tutors will engage in more mental organization of the material and perform more self-explanation as they tutor, leading to learning gains [10,11,16,20,25]....

    [...]

  • ...On the other hand, previous literature has also hypothesized that it is social factors that motivate the tutor effect [3, 7, 11, 15]....

    [...]

  • ...A number of theories have been proposed to explain this effect, including increased motivation to learn the material [23], increased reflection on already learned material [19], and increased effort turning knowledge into coherent, communicable ideas [10,11,29]....

    [...]

Journal ArticleDOI
TL;DR: The model underlying this research program describes how non-screen-and-keyboard-based technologies that listen to children can be used to support their emergent literacy behaviors and have an effect on their subsequent writing skills.

155 citations


""Oh dear stacy!": social interactio..." refers background in this paper

  • ...Given conflicting prior work on whether social relationships can be formed with virtual agents [5,16,17,18] we chose to look at the type of language students used when referring to the agent as a clue to their social stance....

    [...]

  • ...While prior research has shown that children do treat virtual characters similarly to peers in both language use and nonverbal behavior [5], one of the open questions in teachable agent research is whether child tutors are capable of the social motivations described here with a virtual tutee, and whether these social behaviors effect the same tutor learning benefits that can be seen with human peer tutoring....

    [...]

Journal ArticleDOI
TL;DR: This paper explored the influence of offering different instructions to undergraduate students prior to their learning an expository text on evolutionary biology and found that participants were asked to either explain, summarize, or listen to another's explanation or summary of Darwin's theory of evolution through natural selection.
Abstract: This study explored the influence of offering different instructions to undergraduate students prior to their learning an expository text on evolutionary biology. Participants were asked to either explain, summarize, or listen to another's explanation or summary of Darwin's theory of evolution through natural selection. Three conditions were compared: In Condition 1, students were told to study the text because they were going to teach the contents of the text to their partner by either explaining or summarizing (these are referred to as the teach through explanation or summary conditions). In Condition 2, 2 different groups of participants were told to study the identical material on evolution and either explain or summarize the contents of the text aloud to the experimenter following the study period. However, they were only told to do so after they were finished studying for themselves. These are referred to as the explain or summarize to self conditions. In Condition 3, 2 different groups did not read...

154 citations

01 Jan 2001
TL;DR: In this paper, the authors developed computer programs called PALs (Personal A_ssistants for L_earning) in which computers and students alternately coach each other.
Abstract: Our attempts to improve physics instruction have led us to analyze thought processes needed to apply scientific principles to problems—and to recognize that reliable performance requires the basic cognitive functions of deciding, implementing, and assessing. Using a reciprocal-teaching strategy to teach such thought processes explicitly, we have developed computer programs called PALs (P_ersonal A_ssistants for L_earning) in which computers and students alternately coach each other. These computer-implemented tutorials make it practically feasible to provide students with individual guidance and feedback ordinarily unavailable in most courses. We constructed PALs specifically designed to teach the application of Newton’s laws. In a comparative experimental study these computer tutorials were found to be nearly as effective as individual tutoring by expert teachers—and considerably more effective than the instruction provided in a well-taught physics class. Furthermore, almost all of the students using the PALs perceived them as very helpful to their learning. These results suggest that the proposed instructional approach could fruitfully be extended to improve instruction in various practically realistic contexts.

140 citations

01 Jan 2002
TL;DR: Atlas-Andes is a dialogue enhanced model tracing tutor (MTT) integrating the Andes Physics tutoring system with the Atlas tutorial dialogue system, providing Andes with the capability of leading students through directed lines of reasoning that teach basic physics conceptual knowledge, such as Newton’s Laws.
Abstract: Atlas-Andes is a dialogue enhanced model tracing tutor (MTT) integrating the Andes Physics tutoring system (Gertner ~ VanLelm 2000) with the Atlas tutorial dialogue system (Freedman et al. 2000). Andes is a MTT that presents quantitative physics problems to students. Each problem solving action entered by students is highlighted either red or green to indicate whether it. was correct or not. This basic feedback is terlned red-greeu feedback. Furthermore, when students get stuck in the midst of problem solving and request help, Andes provides hint sequences designed to help them achieve the goal of soh’ing the problem as quickly as possible. Atlas provides Andes with the capability of leading students through directed lines of reasoning that teach basic physics conceptual knowledge, such as Newton’s Laws. The purpose of these directed lines of reasoning is to provide a solid foundation in conceptual physics to promote deep learning and to enable students to develop meaningful problem solving strategies. While students in elementary mechanics courses have demonstrated an ability to master the skills required to solve quantitative physics problems, a nmnber of studies have revealed that the same students perform very poorly when faced with qualitative physics problems (Halloun & Hestenes 1985b; 1985a; Hake 1998). Furthermore, the naive conceptions of physics that they bring with them when they begin a formal study of physics do not change significantly by the time they finish their classes (Halloun & Hestenes 1985b). Similarly, MTTs in a wide range of domains have commonly been criticized for failing to encourage deep learning (VanLehn et al. 2000). If students do not reflect upon the hints they are given, but instead simply continue guessing until they perform an action that receives positive feedback, they tend to learn the right actions for the wrong reasons (Aleveu, Koedinger, K~ Cross 1999;

130 citations


""Oh dear stacy!": social interactio..." refers background in this paper

  • ...Instead they tend to answer questions with short keywords, providing no explanation [20, 21]....

    [...]