Dataset for Master's Thesis: AI-powered Chatbots

This dataset contains responses from a survey conducted for a master's thesis at Erasmus University Rotterdam. The survey investigated how consumer perceptions of privacy and trust in interactions with centralized versus decentralized AI-powered chatbots influence customer satisfaction. The survey included a predetermined simulated conversation with an AI-powered chatbot.<b>Purpose of the Study</b>:<br>The main research question addressed in this study is: "How do consumer perceptions of privacy and trust in interactions with centralized versus decentralized AI-powered chatbots influence customer satisfaction?" The study aims to compare the differences in customer satisfaction, privacy concerns, and trust between centralized and decentralized AI-powered chatbots.<br><b>Data Description</b>:<br>This dataset includes responses from 175 participants after data cleaning and removal of incomplete and biased responses. Participants were randomly assigned to one of three groups:<br>Unaware of the chatbot typeInformed they would interact with a centralized chatbotInformed they would interact with a decentralized chatbot<b>Variables</b>:<b>Customer Satisfaction</b>: Measured with Likert scale questions on a 5-point scale from Strongly disagree to Strongly agree.<b>Consumer Privacy Concerns</b>: Measured with Likert scale questions on a 5-point scale from Strongly disagree to Strongly agree.<b>Consumer Trust in AI-Powered Chatbots</b>: Measured with Likert scale questions on a 5-point scale from Strongly disagree to Strongly agree.<b>Consumer AI Familiarity</b>: Measured with Likert scale questions regarding prior usage and understanding of AI technology on a 5-point scale from Strongly disagree to Strongly agree.<b>Demographic Information</b>: Age group, gender, highest education finished, nationality, and occupation.<b>Chatbot Type</b>: Categorical variable with values: 0 for not aware, 1 for aware of interacting with a centralized chatbot, and 2 for aware of interacting with a decentralized chatbot.<b>Usage Notes</b>:<br>The dataset is provided in a XLSX file format and includes all necessary variables for analysis. The dataset can be used to conduct various statistical analyses, including descriptive statistics, hypothesis testing, and regression analysis.

本数据集收录了鹿特丹伊拉斯姆斯大学（Erasmus University Rotterdam）一篇硕士论文所开展的调研的受访者反馈。该调研聚焦于消费者在与集中式及去中心化人工智能驱动聊天机器人（AI-powered chatbot）交互时的隐私感知与信任度，对客户满意度的影响机制。调研过程中预设了一段与人工智能聊天机器人的模拟对话。

#### 研究主题
本研究的核心研究问题为：“消费者在与集中式及去中心化人工智能驱动聊天机器人交互时的隐私感知与信任度，如何影响客户满意度？”本研究旨在对比集中式与去中心化人工智能驱动聊天机器人在客户满意度、隐私顾虑及信任度维度上的差异。

#### 数据概况
本数据集经过数据清洗，剔除无效与存在偏倚的样本后，共包含175名参与者的反馈。参与者被随机分配至以下三组：
1. 不知情聊天机器人类型组
2. 被告知将与集中式聊天机器人交互组
3. 被告知将与去中心化聊天机器人交互组

#### 变量说明
1. **客户满意度**：采用李克特（Likert）5级量表进行测量，评分选项从“完全不同意”至“完全同意”。
2. **消费者隐私顾虑**：采用李克特5级量表进行测量，评分选项从“完全不同意”至“完全同意”。
3. **消费者对人工智能驱动聊天机器人的信任度**：采用李克特5级量表进行测量，评分选项从“完全不同意”至“完全同意”。
4. **消费者人工智能熟悉度**：针对人工智能技术的过往使用情况与认知程度，采用李克特5级量表进行测量，评分选项从“完全不同意”至“完全同意”。
5. **人口统计学信息**：涵盖年龄分组、性别、最高学历、国籍与职业。
6. **聊天机器人类型**：分类变量，取值说明如下：0代表不知情聊天机器人类型，1代表知晓将与集中式聊天机器人交互，2代表知晓将与去中心化聊天机器人交互。

#### 使用说明
本数据集以XLSX文件格式提供，包含所有分析所需的变量，可用于开展各类统计分析，包括描述性统计、假设检验与回归分析。