Uncertainty is an intrinsic feature of real-world information and human judgment. People often communicate uncertainty in everyday language, especially when reporting attitudes and evaluations. However, modern quantitative measurement has popularized precise numerical representations. While precision brings clear benefits such as comparability and computability, when information is inherently vague or imprecise, forcing a single value can limit authentic expression and even move away from the truth itself.

This PhD aims to develop methods that enable end-to-end communication of uncertain information. The research is structured around the following stages:

Uncertainty Elicitation for Authentic Information investigates how interval-valued response formats can better capture inherent uncertainty compared to traditional single-point scales.

Uncertainty Interpretation for Reliable Measurement develops methods for assessing the internal consistency of interval data.

Uncertainty Summarisation for Efficient Overview investigates methods to simplify interval representations into communication-efficient forms while retaining key features.

Uncertainty Presentation for Well-Informed Decision Making extends conventional diagnostic tools to their interval-valued versions, enabling practitioners to identify whether observed differences are genuine or possibly the result of false precision.

To bridge theory and practice, this research also includes the development of open-source software that consolidates the complete workflow, making these methods accessible to researchers and practitioners from different backgrounds. The methods are demonstrated using UK rail passenger survey data, a context where the intangible nature of service and the subjectivity of customer evaluation make uncertainty particularly important to understand and communicate. This research assists rail stakeholders in better understanding customer needs and making well-informed decisions on service prioritisation.