Measuring trust: a text analysis approach to compare, contrast, and select trust questionnaires

Area of Science:

• Psychology
• Human-Computer Interaction
• Information Science

Background:

• Trust is a critical construct in human-human and human-technology interactions.
• Numerous questionnaires exist to measure trust, but their differences are not well understood.
• Systematic comparison of trust measurement tools is needed.

Purpose of the Study:

• To systematically explore differences and common themes across widely used trust questionnaires.
• To provide data-driven guidelines for selecting appropriate trust measurement instruments.
• To develop an interactive tool for exploring the semantic space of trust questionnaires.

Main Methods:

• Conducted a review to identify 46 trust questionnaires across Automation, Humans, and E-commerce domains.
• Utilized GloVe word embeddings to encode questionnaire items and entire questionnaires.
• Employed UMAP for dimensionality reduction and visualization, implemented in a web app.

Main Results:

• Developed a semantic lexicon of trust-related words.
• Provided recommendations for questionnaire selection based on item dispersion and domain/layer composition.
• Created an interactive web application for exploring trust questionnaires.

Conclusions:

• The study offers a novel method for comparing and selecting trust questionnaires.
• Insights into trust measurement can be gained from spoken dialogue or written comments.
• The findings facilitate more informed choices in trust research and application.