Trust Calibration in Information Environments (Dataset, n = 394)

This dataset is part of the Human Clarity Institute’s AI–Human Experience 2026 data series. It examines trust calibration in digital information environments, including perceived reliability of AI-generated content, human intervention thresholds in automated systems, and confidence in digital decision-support outputs. The dataset includes:• validated 1–7 Likert-scale agreement items• structured measures of trust calibration and verification confidence• behavioural indicators of information verification and AI override frequency• multi-select variables stored as canonical semicolon-delimited snake_case tokens• open-text reflections with minimal safe cleaning (trim + newline removal only)• demographic variables across six English-speaking countries• digital life exposure (daily hours online) and AI-tool usage frequency Data were collected on 2026-02-09 via Prolific from adults in the UK, US, Australia, Canada, New Zealand, and Ireland.All data were cleaned, anonymised, and processed under the Human Clarity Institute’s machine-readable dataset protocol, which includes: • canonical snake_case variable naming• validated numeric ranges• standardised multi-select formats• minimal safe text cleaning• full alignment with the accompanying data dictionary• removal of Prolific IDs and timestamps• SHA-256 checksums for all files This dataset contributes to understanding how individuals calibrate trust in AI-mediated information environments, supporting longitudinal tracking of trust stability, intervention thresholds, and verification behaviours as AI systems become increasingly embedded in everyday digital life.本数据集隶属于人类清晰研究院（Human Clarity Institute）2026年人工智能-人类体验数据系列，聚焦数字信息环境中的信任校准问题，涵盖人工智能生成内容的感知可靠性、自动化系统中的人类干预阈值，以及数字决策支持输出的可信度评估。 本数据集包含以下内容： • 经过验证的1-7级李克特量表（Likert-scale）同意度条目 • 信任校准与验证置信度的结构化测量指标 • 信息验证行为与人工覆盖人工智能系统频率的行为指标 • 以标准分号分隔的蛇形命名法（snake_case）Token存储的多分类选择变量 • 仅经过最低限度安全清洗（仅去除首尾空格与换行符）的开放式文本反思数据 • 覆盖六个英语国家的人口统计学变量 • 数字生活暴露量（每日在线时长）与人工智能工具使用频率 本数据集于2026年2月9日通过Prolific平台，面向英国、美国、澳大利亚、加拿大、新西兰与爱尔兰的成年群体收集。所有数据均按照人类清晰研究院的机器可读数据集规范进行清洗、匿名化与处理，该规范包括： • 标准蛇形命名法的变量命名规则 • 经过验证的数值范围设定 • 标准化的多分类选择格式 • 最低限度的安全文本清洗流程 • 与配套数据字典完全对齐 • 移除Prolific平台ID与时间戳 • 所有文件均附带SHA-256校验和 本数据集有助于理解个体如何在人工智能介导的数字信息环境中校准信任，可支持随着人工智能系统日益嵌入日常数字生活时，对信任稳定性、干预阈值与验证行为进行纵向追踪。