Influence of the standard stimulus in magnitude estimation of time intervals

Visually presented time intervals between 100 and 1100 ms were scaled directly. Six different standards within the same range were used. The dependence between the reports and the presented time intervals was found to be linear. Different standards led to estimates lying on different straight lines. The slope of these straight lines decreases upon increasing the standard. Nevertheless the ratio between two neighbouring estimates of each straight line remains invariant with respect to the changes in the standard.