From this, it can be concluded that
at least one of the two types of duration judgments is prone to some form of decisional bias, though no further conclusion can be drawn based on the data presented here. The possible decisional bias operating in
Experiment 1 (which would shift PSE values leftward in a manner indistinguishable from perceptual bias) has already been well described. Here, we propose an additional, previously unconsidered form of decisional bias operating in
Experiment 2. When judging whether the squares are the same or different durations, participants' responses may be affected not only by whether the squares are perceived as the same or different
durations but also by whether the squares are the “same” or “different” in a more abstract sense. More specifically, participants may judge whether the
spatial and temporal dimensions of each square in the square pair are “congruent” or “incongruent.” If the small square is presented for the shorter duration, this square has spatial and temporal dimensions that are the “same” in the sense that both are the “lesser” magnitudes. Conversely, if the small square is presented for the longer duration, this square has spatial and temporal dimensions that are “different.” On congruent trials, both squares have spatial and temporal dimensions that are the “same” in this sense, even though one square is small and short and the other is large and long. Similarly on incongruent trials, both squares have “different” spatial and temporal dimensions. The effect of such a decisional bias would therefore be to
raise the proportion of “same duration” responses for pairs of squares whose spatial and temporal dimensions are congruent (
x-axis values > 0 ms, right half of
x-axis, refer to
Figure 3) and
lower the proportion of “same duration” responses for pairs of squares belonging to the incongruent condition (
x-axis values < 0 ms, left half of
x-axis). This asymmetrical raising and lowering of the response function would translate into a rightward shift of PSE values, equivalent to what would be observed if larger stimuli are
perceived as shorter than smaller stimuli. It is worth noting that, despite this potential source of decisional bias in the equality duration judgment task, this task nevertheless remains an improvement on the comparative duration judgment task for the purposes of testing the claim that larger stimuli are perceived to last longer. This is because the decisional bias proposed to operate in the equality duration judgment task (
Experiment 2) would shift PSE values to the
right, therefore any observed
leftward shift of PSE values in this experiment would provide unambiguous evidence that larger stimuli are perceived to last longer given that a PSE shift in this direction cannot be attributed to decisional bias. This is not the case for the comparative duration judgment task (
Experiment 1) in which a leftward PSE shift could represent either a genuine perceptual effect or a decisional bias.