Biomedical Engineering Reference
In-Depth Information
of conditions that varied from easily passable at slow walking speeds to impassable
at fast walking speeds. In one condition, which I will refer to as the Move condition,
subjects began walking from a designated home location at the same time that a pair
of cylindrical obstacles began converging toward a point along their future path. The
cylinders disappeared 1 s after they began moving and subjects were instructed to
press one of two buttons on a handheld remote mouse to indicate whether they could
have safely passed through the gap before it closed. In another condition, which I
will refer to as the Delayed Move condition, subjects waited at the home location
for 1 s after the cylinders began moving. After 1 s, an auditory “go” signal was pre-
sented instructing participants to begin walking. At the same time as the go signal,
the cylinders disappeared and subjects were instructed to judge whether they could
have safely passed through the gap before it closed.
This manipulation allowed [
6
] to test the prediction of the information-based
hypothesis that subjects can accurately perceive action-scaled affordances only while
moving. In the Move condition, subjects were allowed to move for 1 s while viewing
the cylinders before making a judgment. In the Delayed Move condition, subjects
were stationary for the entire 1 s during which the moving cylinders were visible and
did not begin moving until the cylinders disappeared. Therefore, if the information-
based hypothesis is correct, judgments about the passability of the gap should be
accurate in the Move condition but not in the Delayed Move condition.
To measure the accuracy with which subjects perceived passability in the Move
and Delayed Move conditions, judgments were compared with behavior on two other
sets of trials in which the obstacles remained visible and subjects actually attempted to
pass through the gap. That is, for both the Move and Delayed Move conditions, there
was a corresponding set of trials in which the cylinder remained visible beyond 1 s and
subjects attempted to walk through the gap. When judgments were compared to actual
behavior, there was no evidence that subjects either overestimated or underestimated
their ability to safely pass through the gap. That is, subjects tended to judge that
the gap was passable in conditions in which they were able to pass through the gap
and impassable in conditions in which they were unable to pass through the gap.
Furthermore, this was true in both the Move and Delayed Move conditions. Thus,
regardless of whether subjects made the judgment while moving or while stationary,
they were able to accurately perceive passability. Such findings do not support the
information-based account.
1
To summarize, the information-based approach explains how action-scaled affor-
dances, such as the passability of a shrinking gap, could be perceived by relying
entirely on visual information without appealing to stored knowledge of move-
ment capabilities. However, because the visual information upon which actors are
thought to rely is only available “on the fly,” the information-based approach predicts
that action-scaled affordances cannot be accurately perceived while stationary [
25
].
1
In an earlier study, Oudejans et al. [
25
] found that judgments were more accurate when subjects
were allowed to move. However, Fajen et al. [
6
] pointed out several methodological problems that
explain their findings and showed that once these problems were corrected, judgments were equally
accurate regardless of whether subjects were stationary or moving.
