A comparison of the augmented bar display and the object display

Two experiments were performed to test the hypothesis that object displays facilitate integration task performance because they contain information that is not readily accessible in standard bar displays. The first experiment demonstrated that when the extra information present in an object display is added to a bar display, as in the augmented bar display, integration task performance can be improved in the bar display. The second experiment shows that integration task performance is poor with an object display when the extra information present in the object display is not relevant to the integration task. If the decision statistic can be specified at the time of display design, an augmented bar display may be appropriate since it offers good performance for both separate and integral tasks; if the decision statistic cannot be specified at the time of display design, the use of an object display may hinder performance if the decision statistic and emergent feature are not correlated.