A Structural Metric Model to Predict the Complexity of Web Interfaces

• Main Author: Attaallah, Abdulaziz Ahmad
• Format: Others
• Published: North Dakota State University 2017
• Online Access: http://hdl.handle.net/10365/25918

The complexity of web pages has been widely investigated. Many experimental studies used several metrics to measure certain aspects of the users, tasks or GUIs. In this research, we focusing on the visual structure of web pages and how different users look at them regarding complexity. Several important measures and design elements have rarely been addressed together to study the complex nature of the visual structure. Therefore, we promoted a metric model to clarify this issue by conducting several experiments on groups of participants and using several websites from different genres. The goal is to form a metric model that can assist developers to measure more precisely the complexity of web interfaces under development. From the first experiment, we could draw the guidelines of the major entities in the metric model, and the focus was on two most important aspects of the web interfaces, which are the structural factors and elements. Thus, four main factors and three main elements were more representatives to the concept of complexity. The four factors are size, density, grouping and alignment, and the three elements are text graphics and links. Based on them we developed a structural metric model that relates these factors and elements together, and the results of the metric model are compared to the web interface users’ ratings by using statistical analysis to predict the overall complexity of web interfaces. The results of that study are very promising where they show our metric model is capable of predicting the complex nature of web interfaces with high confidence.