Two photos show Sue wearing a rain coat and holding a long white cane while standing at the corner of two 2-lane streets. The first photo shows the street she is facing continue to her left and then disappear around a corner, the second photo shows the street to her right continue straight for at least a block, with a bus stop sign beside her.

Crossing in a Situation of Uncertainty -- What Happened?

Every day for 15 years, Sue Etters got off the bus on her way home from work and crossed Beaver Avenue where there is no stop sign or signal. The crossing there is just two lanes, each 14 feet wide. To the right, the street goes straight for more than a block and to the left, the road disappears around a bend about 150 feet away (see photos to the right). Pedestrians have the right of way where she was crossing because according to Pennsylvania law, it is an unmarked crosswalk.

Sue said she always waited until it was quiet before she started to cross, as she had been taught to do in the 1960's. However, sometimes her guide dog stopped her when she was about to walk into the path of a car that she hadn't heard when she started to cross.

Starting July 2, 2009 Sue crossed with a cane, and was scheduled to get her next guide dog the following January. However at 6:42 PM December 16, 2009, she was hit by a car and severely injured. She doesn't remember anything about the crash, but according to the police report, the roads were dry, and she was hit in the middle of the first lane by a driver coming from around the bend on her left.

Preliminary observations of crossing where Sue Etters was hit:
In October, 2010 I went to Pittsburgh and observed the crossing and discussed it with Sue. The roads were wet so I was unable to analyze the situation that probably existed when she was hit. However, out of curiosity, after Sue left I used the TMAD during a 45-minute observation to analyze whether it was a Situation of Uncertainty for someone with hearing like mine. This is what I observed:

Time needed for crossing:
Number of vehicles observed approaching from left when quiet:
Range for times of detection-to-arrival from the left:
Conclusions for crossing first half of the street in these conditions (the road was wet):
Considering "quiet"?

Observations from second visit:

Detection-to-arrival time
for vehicles heard when quiet

left vehicles

average=7.6

right vehicles

average=9.9

6.0 seconds

6.6 seconds

6.8 seconds

7.0 seconds

7.8 seconds

8.0 seconds

9.0 seconds

9.8 seconds

8.0 seconds

8.5 seconds

8.6 seconds

9.3 seconds

9.9 seconds

11.3 seconds

12.6 seconds

A year later, on Friday September 16, 2011, I had an opportunity to go back to Pittsburgh and evaluate Sue's crossing again when the roads are dry, as they were on the day that she was hit. I arrived at the intersection and started observing at 4:30. In the next 25 minutes, I had gotten 18 periods of complete "quiet."

After 5 o'clock, however, there were no more periods of quiet. Sounds would die down and become very faint, but before they completely disappeared I could hear the sound of another approaching vehicle. I recorded the detection-to-arrival (warning times) for vehicles heard when it was completely quiet as well as those that were heard when there were still some faint sounds in the distance.

When it was completely quiet I heard vehicles from the left 6-10 seconds away (average 7.6) and from the right 8-13 seconds away (average about 10 -- see chart to the right). There was a Situation of Uncertainty for vehicles from the right (she needed 9 seconds to cross and I couldn't hear some of them until they were 8 seconds away).

But when it was completely quiet, all the vehicles from the left were heard with enough warning. The shortest warning time (6 seconds) was only 1.5 seconds longer than crossing time (4.5 seconds) but the range from the shortest warning time to the longest was only 4 seconds, so it's highly unlikely that the warning time of any approaching vehicles would be much shorter than the crossing time.

So what would explain that when she thought it was quiet and she started to cross, there was an approaching vehicle so close to her that it hit her?

Detection-to-arrival time
with slight masking sound

left vehicles

average-5.6

right vehicles

average=6.5

3.4 secs

4.2 secs

4.4 secs

4.8 secs

4.9 secs

5.0 secs

5.1 secs

7.2 secs

7.6 secs

8.0 secs

9.8 secs

4.2 seconds

5.0 seconds

5.6 seconds

6.0 seconds

6.3 seconds

6.4 seconds

6.7 seconds

7.0 seconds

7.1 seconds

7.7 seconds

8.3 seconds

9.3 seconds

The answer may be the fact that the warning time (detection-to-arrival time) for approaching vehicles was dramatically reduced whenever there was even slight masking sounds, as seen in the chart to the right. This was revealed dramatically when I compared the warning times for vehicles that were heard when quiet with those that were heard when there was a slight masking sound (see chart to the right).

Vehicles were heard from the left an average of 7.6 seconds when quiet but only 5.6 seconds when there was a slight masking sound. The shortest warning time from the left when quiet was 6 seconds, but when there was a slight masking sound one vehicles was not heard heard until it was only 3.4 seconds away.

I then arranged for Sue to come to the site, and I asked her to report when she thought it was quiet enough to cross. Indeed, she reported it was "quiet" when there were slight masking sounds present. This may explain why she was unable to hear the car and thought it was clear to cross even though it was so close. Sue didn't realize that the only way for her to be confident that it is clear to cross is to wait until all masking noises (including the sound of the bus that she had just gotten off) had completely disappeared. Since I observed no such opportunities of quiet after 5:00, the time when she routinely came home and crossed there, she would have benefitted from analyzing the risk of crossing and deciding if the risk was acceptable and if not, consider alternatives.

Our students need to be able to recognize Situations of Uncertainty and, in Situations of Confidence, recognize the effect of masking sounds and realize how quiet it must become before they can hear the approaching vehicles well enough to be confident that it is clear to cross.

Return to Situations of Uncertainty . . . does it really matter?
Return to home page