Hazard anticipation of young novice drivers - SWOV

animation clips were made, each lasting about thirty-five seconds. All videos
were 'taken' from the driver's perspective. Animation was used because it is
easier to stage scenes in animation videos than in real life videos. In none of
the clips the hazards materialized. This means that no car suddenly emerged
from behind an object that blocked the view and no other road users
performed risky actions.
An interactive pointing system (with a computer mouse) was applied.
At certain moments in each clip, the clip was paused for 5 s. These short
pauses with a frozen traffic situation on the screen were necessary to allow
participants to make mouse clicks. During these pauses participants could
click on covert latent hazards and on overt latent hazards. They also could
click on small rectangles located at the left and right side of the screen. With
a mouse click on the right rectangle participants could indicate that they
would have looked to the right in search of a hazard if this would have been
possible. With a mouse click on the left rectangle participants could indicate
the same for hazards to the left. In each pause only the first three clicks,
including the clicks on the rectangles were recorded. To ensure that
participants concentrated on the task and did not randomly click as many
times as possible, participants were told that clicks on irrelevant spots would
lower their score. Not all pauses contained three latent hazards and in some
pauses, there were no latent hazards at all. A disadvantage of pauses is that
participants can reorient themselves (McGowan & Banbury, 2004). This is to
say, the very fact that a video clip pauses, could be a cue for participants to
start searching for latent hazards they had not noticed while the video was
still running. The pause itself, in contrast to driving in the real world, also
facilitates participants with time to search for these possible latent hazards.
This was another reason to instruct participants in advance that false clicks
would lower their score and to keep the pauses short (5 s).
Just as in the hazard detection and recognition task of Chapter 4, before the
start of each video clip a plan view was presented of the manoeuvre the car
would make in the video clip. The number of pauses per video clip differed
and ranged between three and six per clip. When a pause started, a countdown
timer was visible on the bottom of the screen, indicating the time left
for pointing and clicking. On spots where a click was made a green cross
appeared. However, after three clicks in a pause no more crosses appeared
when a click was made. In contrast to video clips used in Chapter 4, no part
of a dashboard and a steering wheel was visible. See for an example of a
pause Figure 5.3.
175