Empirical Lab Repository

Title: Evaluating the efficiency of a user interface: Comparing collected timings with predictions from the Keystroke-Level Model

Author: Craig Miller, DePaul University, cmiller@cs.depaul.edu

Possible Courses: Human-Computer Interaction, User-interface design

Empirical Concepts Used: confidence intervals, sources of error in collecting timings of human users, limitations of theoretical models

Computer Science Concepts Used: User interface efficiency, Card Moran and Newell's Keystroke-level model, timed usability testing

Summary: Students perform two methods for evaluating the efficiency of a user interface. Here efficiency is defined as the time needed for a user to complete a core task using the interface. The first method is Card, Moran and Newell's Keystroke-Level Model (KLM). Students apply the KLM by analyzing the keystrokes needed to complete the task. The KLM provides time constants and rules that produce a prediction for expert usage time. The second method involves empirical data collection. Students collect times of classmates completing the task. Students calculate averages and confidence intervals of the collected times and compare them to the theoretical prediction of the KLM. This assignment promotes discussion of possible sources of error in both theoretical models and empirical data collection.

• Lab Handout (no programming)
  
  
• Instructor Materials
• Assignment Evaluation Form