“The Whys, How Tos, and Pitfalls of User Studies” by Sundstedt, Whitton and Bloj

  • ©Veronica Sundstedt, Mary C. Whitton, and Marina Bloj


Entry Number: 05


    The Whys, How Tos, and Pitfalls of User Studies



    Members of the SIGGRAPH community are both consumers and producers of algorithms that make images and techniques that let us interact with visual applications. This course explains the essential role of user studies in insuring that algorithms, products, and content are effective for their intended purposes. The course introduces user studies through real examples and case studies that highlight good practices and warn of mistakes that can compromise evaluation. The cases are chosen to demonstrate the range of the application of user studies in computer graphics (for example, in developing better algorithms and in evaluating images and interaction techniques. The first part of the course summarizes the different types of studies that are appropriate at different times during development of a user-interface technique. The second part uses examples from the evaluation of graphics to illustrate the different techniques that can be used to assess perceived image quality and preference. It includes some well-known techniques such as ranking and introduces novel concepts such as maximum-likelihood difference scaling.
    The third part reviews how eye tracking can support user studies. Eye tracking is useful as an interaction device in virtual environments, but it can also be a helpful tool in usability testing and evaluation of algorithms and techniques. It can collect and analyze additional data that can not be measured explicitly using questionnaires.

    Level: Introductory