S. C. 3/28/2008 (Schedule is Changed)

Time: Fri. 3/28, 1pm
Location: SZB 556
Speaker: Sheng-Cheng (Hans) Huang

Genre: Individual Studies

Topic and Abstract: Menu-Navigating Behavior (MNB): A model of quantifiable and optimal user behaviors in finding paths within a limited information architecture

Today, when a user uses a computer to perform given tasks, his or her behaviors are mostly dominated by motor skills of operating hardware such as the mouse and the keyboard and by cognitive mechanisms of making either perceptive or integrated decisions to select correct items such as icons and labels on the screen. In this study, we will boldly characterize human-computer interaction (HCI) as a series of tasks of menu-navigating behaviors, which is to deductively define menu-navigating as a iterative behavior of choosing the correct visual stimulus among others by perceptive and cognitive mechanisms in order to reach the final destination of a target item.

Deriving from this definition, we assume that for a user to navigate a menu system and to successfully find the target information is analogous to the behavior of finding the correct path that is predefined by the designer of the system. Therefore, we are interested in finding the facts of how users make correct decisions that leads to form the correct path and vice versa. Also, based on the facts that we expect to identify and discover, we hope to develop a mathematic model of this particular HCI based on the statistical probability of user behaviors that will provide reasonable parameters to help us design better interfaces and predict user performance in the future.

The purpose of this study is to analyze the user behaviors in menu-navigating tasks, and to examine whether the current psychometric functions of Two Alternative Forced Choice (2AFC) task, Theory of Signal Detectability (TSD) and Receiver Operating Characteristic (ROC) can be applied to understand this particular phenomenon. I also propose to develop new or adjust current psychometric functions based on the performance data to be collected in order to answer the question whether users always apply rational rules (e.g. "ideal" Bayesian decision rule) for probability judgment or decision-making in menu-navigating tasks.


Attendance: 10