Impact of technological support on the workload of software prototyping

Suleri, Sarah; Jarke, Matthias (Thesis advisor); Prinz, Wolfgang (Thesis advisor); Schröder, Ulrich J. (Thesis advisor)

Aachen : RWTH Aachen University (2021)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2021

Abstract

Prototyping is a broadly utilized iterative technique for brainstorming, communicating, and evaluating user interface (UI) designs. This research aims to analyze this process from three aspects: traditional UI prototyping, rapid prototyping, and prototyping for accessibility. We propose three novel approaches and realize them by introducing three artifacts: 1) Eve, a sketch-based prototyping workbench that supports automation of transforming low fidelity prototypes to higher fidelities, 2) Kiwi, a UI design pattern and guidelines library to support UI design pattern-driven prototyping, 3) Personify, a persona-based UI design guidelines library for accessible UI prototyping. We evaluate the usability of these artifacts, and the results indicate good usability and learnability. Furthermore, we use NASA-TLX to study the impact of using these three novel approaches on the subjective workload experienced by the designers during the software prototyping process. Our workload analysis reveals that, unlike the traditional prototyping approach, Eve’s comprehensive support eliminates the need for switching between various prototyping tools while progressing through the low, medium, and high fidelity prototypes. Consequently, there is a significant decrease in subjective workload experienced by designers using the comprehensive approach offered by Eve. Also, there is a significant reduction in mental demand, temporal demand, effort, and five times increase in the overall perceived performance using the comprehensive approach (Eve). Similarly, the subjective workload experienced by designers using the pattern-driven approach using Kiwi is significantly less than the workload experienced using the traditional approach of rapid prototyping. Specifically, there is a significant decrease in physical demand and effort of rapid prototyping while using the pattern-driven approach. Lastly, the subjective workload experienced by UI/UX designers using the persona-driven approach offered by Personify is significantly less than the workload experienced using the traditional approach of prototyping for accessibility. Specifically, there is a significant decrease in mental demand and effort of prototyping accessible UIs while using Personify. This work aims to extend prior work on UI prototyping and is broadly applicable to understand the impact of using deep learning, UI design patterns, and personas on the workload of UI prototyping.