Definition and development of a measurement instrument for compellingness in human computer interaction
Overly compelling displays may cause users to under or overestimate the validity of data that is presented, leading to faulty decision making, distractions and missed information. However, no measure currently exists to determine the level of compellingness of an interface. The goal of this research was to develop an empirically determined measurement instrument of the compellingness of an interface. Literature review and a semantics survey were used to develop a pool of items that relate or contribute to compellingness, and two expert reviews of the list resulted in 28 potential questions. These 28 questions were fielded in study with a map-based task. Exploratory Factor Analysis and Cronbach’s Alpha were used on the results to eliminate questions, identify factor groupings, and quantify the amount each question loaded on the factor groupings. That analysis resulted in a final compellingness survey with 22 questions across six sub-factors and a final Cronbach’s Alpha value of 0.92. Additionally, the survey is organized into three factors of compellingness: human, computer, and interaction, resulting in a two-level survey. An empirically-based measure of compellingness can be used in evaluations of human factors issues in domains such as aviation, weather, and game design. Understanding the underlying aspects of compellingness in an interface will enable researchers to understand the interaction between compellingness and other human factors issues such as trust, attention allocation, information quality, performance, error, and workload.