Project topics and materials
Get Complete Project Material File(s) Now! »
Social Issues of Human-Robot Interaction
Companion robots will be more and more part of our daily lives in the coming years, and having long-term interactions with them can have both positive and negative effects on their users.
With robots being more and more around people, situations where social interactions have an effect (positive or negative) can appear more frequently. For this reason the robots need to be capable of adapting to the user so as to improve the interaction and the user’s task performance. This chapter presents an experiment focused on social facilitation, which is the evaluation of performance in the presence of others, as described in the social psychology literature. Our scenario is a memory game played with cards against a Nao robot. The robot has a framework combining an emotional system based on the OCC Model, and an episodic memory mechanism.
Social Facilitation
The aspect discussed in the work presented in this chapter is called Social Facilitation, which is a widely studied (Michaels et al., 1982) (Uziel, 2007) psychology paradigm introduced by (Zajonc et al., 1965), that states that individuals get a better performance on easy tasks if they are in the presence of others than doing the same task alone, but their performance is worse in complex tasks. Very little work in social robotics (Wechsung et al., 2014) (Riether et al., 2012) or virtual characters (Park and Catrambone, 2007) has focused on Social Facilitation. The authors in Riether et al., 2012 presented a study that compared the task performance of 106 participants on easy and complex cognitive and motor tasks across three presence groups (alone, human presence, and robot presence). They found evidence that confirms the theory of Social Facilitation, but they focused on the mere presence of the robot.
This chapter presents an experiment where the social facilitation effect in Human-Robot Interaction is investigated. The scenario involves a memory card game in which the robot is the opponent of the human player, and it can take two roles: it can encourage or judge the human-user, depending on the game mode.
Choosing an Emotion Model for the Robot
In order to provide the robot with a mechanism capable of generating a natural behavior, we included in our framework an emotional system. The literature indicates the existence of mainly two broad categories of emotion models (cognitive and non-cognitive), which were reviewed in Chapter 2. A non-cognitive model has the advantage of providing a fast response to the robot, because it is more reactive. Nevertheless, we opted for a cognitive model because we wanted to have the capability of adapting the behavior of the robot during the interactions with its users, and this capability needs a memory and mechanism of reasoning about past events. Among the cognitive models of emotion, there are many existing models, and some of them have been implemented on robotic systems, e.g., (Gadanho and Hallam, 2001), (Fellous, 2004), (Dodd and Gutierrez, 2005). We decided to use the OCC Model (Ortony, Clore, and Collins, 1990) because it has many desirable characteristics for us, such as a wide range of emotions, three categories based on objects, agents, and events, where the category of events include variables about past events to generate the emotions. Also, it has partially been implemented before on virtual animated characters (André et al., 2000b) and robotics (Kröse et al., 2003).
The OCC Model Ortony, Clore, and Collins, 1990 is based on 4 global variables and 12 local variables, each variable depending on both physical and psychological factors. The model has 22 emotions that are divided in three categories: Aspect of Objects, Action of Agents, and Consequences of Events. In this work, we focused on the category of Action of Agents for its relation with the social facilitation effect.
Table of contents :
Declaration of Authorship
Abstract Fr
Abstract
Abstract Spa
Acknowledgements
1 Introduction
1.1 Context
1.2 Proposed Work
2 Theoretical Foundations
2.1 Emotion Theories
2.1.1 Non-Cognitive Theories
2.1.2 Cognitive Theories
OCC Model
2.2 Adaptation and Memory
Episodic Memory
2.3 Behavior Theories
2.3.1 Personality Trait Theories
Big Five Personality Trait
2.3.2 Regulatory Focus Theory
2.4 Conclusion
3 Integrating a Framework for Human-Robot Interaction
3.1 Social Issues of Human-Robot Interaction
3.1.1 Social Facilitation
3.2 Choosing an Emotion Model for the Robot
3.2.1 Synthesis of Emotions
3.3 Giving an Episodic-Like Memory to the Robot
3.4 Testing the Framework in a Game-Like Scenario
3.4.1 Hypothesis
3.4.2 Robot Behaviors
3.4.3 Methodology
3.4.4 Results and Discussion
3.5 Contributions and Conclusion
4 Need of a User Pattern for a Personalized Interaction
4.1 Choosing a Personality Model
4.2 Differences between Introverts and Extroverts
4.3 Differences between People with High Conscientiousness and People with Low Conscientiousness
4.4 Methods
4.4.1 Case Study: Office-Like Scenario with a Robot Giving Reminders
4.4.2 Hypotheses
4.4.3 Robot Behavior
4.4.4 Pre-experiment Questionnaire
4.4.5 Conditions
4.4.6 Post-experiment Questionnaire
4.5 Results and Discussion
4.6 Contributions and Conclusion
5 Effects of Stress and Personality in HRI using Multimedia Learning
5.1 Robots Used for Teaching
5.2 Multimedia Learning
5.3 Case Study: Teaching Nutrition by using Multimedia with a Robot
5.3.1 Multimedia Learning Scenario
5.3.2 Multimedia design
5.3.3 Conditions
5.3.4 Hypotheses
5.3.5 Post-experiment Questionnaires
5.3.6 Image Analysis
5.3.7 Stress Analysis
5.4 Results and Discussion
5.4.1 Hypothesis 1: Human voice
5.4.2 Hypothesis 2: Robot embodiment and Learning
5.4.3 Hypothesis 3: Robot embodiment and Stress
5.4.4 Hypothesis 4: Stress, and time and test score
5.4.5 Hypothesis 5: Neuroticism
5.4.6 Questionnaire for feedback
5.4.7 Discussion
5.5 Contributions and Conclusion
6 Adapting a Robot Behavior to Personality Preferences
6.1 Choosing the Features to Create the Model
6.2 Learning a Model of Users Preferences of Robot Behavior
6.2.1 Configuring the Episodic Memory
Episodes with specific information
Episodes with personality/gender information
6.2.2 Adapting the OCC Model
Emotion for episodes with specific information
Emotion for episodes with personality/gender information
6.3 Case Study: Learning Users’ Preferences and using them with New Users
6.3.1 Hypothesis
6.3.2 Robot Capabilities
6.4 Methods
6.4.1 Training the robot
6.4.2 Pre-experiment Questionnaire
6.4.3 Post-Experiment Questionnaire – Training Phase
6.4.4 Testing of the model
6.5 Results and Discussion
Training the robot
Testing the model
6.6 Contributions and Conclusion
7 Analysis of Relation Between User Performance and the Regulatory Focus Theory
7.1 Using Regulatory Focus Theory to improve User Performance
7.1.1 Hypotheses
7.2 Case Study: Stroop Test with Robot Instructions Based on the Regulatory Focus Theory
7.3 Methods
7.3.1 Conditions
7.3.2 Regulatory Focus Questionnaire – Proverb Form
7.3.3 Measures
7.4 Results and Discussion
Hypothesis 1
Hypothesis 2
Hypothesis 3
7.5 Contributions and Conclusion
8 Evaluating Different Robot Behaviors Based on the Regulatory Focus Theory
8.1 Regulatory Focus Theory
8.1.1 Regulatory Focus and Regulatory Fit
8.1.2 Regulatory Fit and Negotiation in HRI
8.2 Case Study: Negotiating with a Robot with Behavior based on the Regulatory Focus Theory
8.2.1 Hypothesis
8.2.2 Negotiation game Scenario
8.2.3 Robot Platform
8.2.4 Robot Speech
8.2.5 Robot Speech Recognition
8.2.6 Conditions
Control Condition
Promotion Based Robot Behavior Condition
Prevention Based Robot Behavior Condition
8.2.7 Regulatory Focus Questionnaire – Proverb Form
8.2.8 Measures
8.3 Results and Discussion
8.3.1 Hypothesis 1
8.3.2 Hypothesis 2
8.3.3 Hypothesis 3
8.3.4 Godspeed Questionnaire
8.4 Contributions and Conclusion
9 Adapting Robot Behavior using Regulatory Focus Theory and User Physiological and Task-Performace Information
9.1 Study Case: Adapting Robot Behavior to User Task Performance and Stress in a Game-Like Scenario
9.1.1 Game-like Scenario
9.1.2 Robot behavior
9.1.3 Behavior Adaptation System
9.1.4 Experimental Design Setup
9.1.5 Hypothesis
9.1.6 Training the robot
9.1.7 Testing of the model
9.1.8 Measures
9.1.9 Training the robot
9.1.10 Testing the model
Promotion Participants
Prevention Participants
9.1.11 Godspeed Questionnaire Results
9.2 Contributions and Conclusion
10 Conclusion
10.1 My Contributions
10.2 Limits and FutureWork
A Evaluation Documents
A.1 Big Five Questionnaire
A.2 Regulatory Focus Questionnaire – Proverb Form English
A.3 Godspeed Questionnaire – Sections 1-3
A.4 Game-Like « Find the pair » Scenario Post-Questionnaire
A.5 Office-Like Scenario Post-Questionnaires
A.6 Multimedia Scenario Post-Questionnaire
A.7 Learning Users’ Preferences Post-Questionnaire – Training
A.8 Learning Users’ Preferences Post-Questionnaire – Testing
B Publications List
Bibliography
