Mock-up or eye movement recording : big payback from a 'discount' method for multimedia interface design
Matthias Rauterberg · TU/e Research Portal · 1995
We carried out two different empirical investigations with exactly the same multimedia screens: (1.) an experiment with measurements of eye movements (8 men, 26±3 years), and (2.) a mock-up study (17 women , 34±6 years). From both investigations we got two different kinds of data: (a) the unconscious eye movements as a parameter of the users' attention focus during an interactive task solving process with a multimedia information system [Rauterberg, 1995]; (b) the conscious and intentional users' choices looking on a static screen dump. We used the eye tracking system (ETS) of DORNIER Inc. The subject sat in a normal distance (30-50) in front of a computer screen (17) without any contact to the eye recording measurement unit. In the mock-up study the subjects were instructed to make a virtual mouse click at each screen area of interest. This was done by marking each screen dump with maximal three numbered crosses (first, second, and third choice). Analysing the recorded eye movements we have to differentiate two situations: (1) the user comes to a screen, (2) the user looks around on the screen. In case (1) a user tries to orient himself. After this orientation phase (2) he is looking for the task related information. So, we differentiate between the first fixation (measure: E first) and all fixations (measure: E total). In the mock-up study we analysed the first three different choices (measure: M first, M second, M third) and all choices for each screen (measure: M total = M first + M second + M third). We calculated the correlation coefficients among these six measures for each screen separately and over all four screens (see Table 1). All correlations of the condition 'screens 1-4' (last column in Table 1) are significant. The main result of this methodological investigation is the significant correlation 'E total x M total' (r=.727; p≤.001).