Both the line graph and dot chart, shown at the top of Figure 2, use the vertical position of points from a common baseline to represent Variable 1 (left) and Variable 2 (right).
Contrary to predictions, one object display--the line graph--was associated with more errors than its separable counterpart, the dot chart. Mean object integration effect sizes for homogeneous displays were r = .27 for RTs and r = .11 for errors.
Neither of these options may have been easier than performing the two height judgments with the dot chart. The dots may have made exact position (height) judgments easier because participants could use the vertical line segments, which extended from the dots to their zero points, as a reference scale.