Risk Attitude Reversals in Drivers Route Choice When Range of Travel Time Information Is Provided

1
Risk Attitude Reversals in Drivers Route Choice
When Range of Travel Time Information Is Provided

• Megan Englert
• Tim Leser

2
Purpose

• The goal of this work is to investigate the
  influence of range of travel time on drivers
  risk attitude, and whether a model can account
  for this influence.

3
Background Support

• Katsikopoulos et al. (2000) found that risk
  attitude and route choices were heavily
  influenced by whether or not the route choice
  belonged to the domain of gains or losses.

4
Theoretical Basis

• The basis for analyzing this model is whether or
  not it can predict for which route choice
  scenarios drivers are risk averse or risk
  seeking.
• This idea comes from the language of psychology
  of choice.

5
Practical Contributions

• The efficiency of the transportation networks
  can be improved.
• Support drivers decisions.

6
Theoretical Contributions

• Empirically demonstrates the influence of the
  range of travel times on drivers risk attitude.
  
• A simple model that predicts drivers risk
  attitudes, given route choices

7
Methods

• The study was performed on two routes (I-93
  Route 28). Both of these routes lead to
  downtown Boston.
• Subjects (15 men, 15 women) were recruited from
  the University of Massachusetts and the local
  area.

8
Methods

• A driving simulator was used to perform the
  experiment.
• Subjects were instructed to drive in the right
  lane, and follow a lead vehicle.
• The lead vehicle was used to control the
  subjects speed.

9
Methods

• Subjects were given a reference route (I-93)
  with range of travel time c, and an alternative
  route (Route 28) with a range of travel times
  r.
• 6 virtual traffic databases each containing 1 of
  6 VMS (Variable Message Sign) were used to create
  36 driving scenarios.

10
Methods

• Information of the ranges of travel times for
  each route was displayed in on the VMS on the
  right side of the road
• Subjects spoke aloud their route choice, and the
  experimenter recorded it

11
Analyses (figure 2)
12
Analysis (figure 3)
13
Analyses (figure 4)
14
Analyses

• The data were analyzed using a 2x3x6 ANOVA, with
  repeated measures.

15
Results

• Drivers diverted significantly less often as the
  alternative expectation increased.
• The effect of alternative range was also
  significant. This depended on the level of
  alternative expectation.

16
Results

• The results imply that drivers are risk seeking
  when (gains) and risk averse (losses) when the
  reference range of travel times is larger than
  the alternative range.

17
Future Work

• Perform these tests in other cities, without
  letting participants know what the study is.
  Possible by using GPS devices.
• This work was performed when the subjects had
  two choices, try to model n route choices.

18
Discussion

• Questions
• Thoughts