Travel time uncertainty and ambiguity attitudes

Last registered on May 03, 2023

View Trial History

Pre-Trial

Trial Information

General Information

Title

Travel time uncertainty and ambiguity attitudes

RCT ID

AEARCTR-0011248

Initial registration date

April 28, 2023

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

May 03, 2023, 4:25 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

United Kingdom of Great Britain and Northern Ireland

Region

Primary Investigator

Name

Kexin Geng

Affiliation

Vrije Universiteit Amsterdam

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Xiao Yu

PI Affiliation

Vrije Universiteit Amsterdam

Contact Investigator

PI Name

Erik Verhoef

PI Affiliation

Vrije Universiteit Amsterdam

Contact Investigator

Additional Trial Information

Status

Completed

Start date

2023-04-28

End date

2023-05-02

Keywords

Behavior

Additional Keywords

online experiment, travel time uncertainty, ambiguity attitudes, willingness to pay, car, public transport

JEL code(s)

R41, D81, D83

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

There is growing attention to the complex pattern of uncertainty attitudes in the traffic context. Existing studies often assume that travellers treat travel time uncertainty as risk, i.e., outcomes with a known probability distribution. However, travellers often face ambiguous travel times, i.e., with an unknown probability distribution. In this study, we develop a novel framework of uncertainty attitudes in the transport context to investigate decision-making under uncertainty for commuting travel, which are day-to-day types of repeated situations. Specifically, we focus on the role of ambiguity attitudes in mode choice. Furthermore, we aim to establish the potential link between source preferences and mode preferences. Our online experiment design consists of a baseline group and two treated groups: one with additional information from historical travel time and another considering a commute trip from participants' real life.

External Link(s)

Registration Citation

Citation

Geng, Kexin, Erik Verhoef and Xiao Yu. 2023. "Travel time uncertainty and ambiguity attitudes." AEA RCT Registry. May 03. https://doi.org/10.1257/rct.11248-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

1. Control group: The benchmark treatment, uses a certain OD pair for all respondents.
2. Information group: Give respondents additional information about historical travel time in the last week.
3. Individualized group: use individualized OD pair for each respondent.

Intervention Start Date

2023-04-28

Intervention End Date

2023-05-01

Primary Outcomes

Primary Outcomes (end points)

a: ambiguity aversion
g: (ambiguity generated) insensitivity index
b_i (i=1,2,3): ambiguity neutral probabilities (i.e., beliefs of each state-of-world)

Primary Outcomes (explanation)

For each travel mode (make the trip by car or by public transport), respondents need to answer six questions. Three of them show three independent travel time intervals, which together form a complete set of possible travel times. And the other three questions show their complementary time intervals, separately. Between every pair of Option 1 and Option 2, we elicit indifference point/matching probability: p1, p2, p3, p12, p23, p13, for car and public transport respectively.

p_s=(p_1+p_2+p_3)/3 denote the average single-event matching probability.
p_c=(p_12+p_23+p_13)/3 denote the average composite-event matching probability.

The ambiguity aversion index is: a=1-p_s-p_c
The (ambiguity-generated)-insensitivity index is: g=3×[1/3-(p_c-p_s )]
The ambiguity-neutral probabilities (Beliefs): b_1=(p_12+p_13-p_2-p_3)/(6(p_c-p_s)) , b_2=(p_12+p_23-p_1-p_3)/(6(p_c-p_s)) , b_3=(p_23+p_13-p_1-p_2)/(6(p_c-p_s))

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

In the questionnaire, participants in the treatments groups will be asked to play the lotteries, a) complemented with information about the travel time of the past 4 trips in the last week by researchers, or b) about the travel time of the researchers if they travel similarly as participants’ own regular morning trip.

Experimental Design Details

Users are randomly assigned to either the control or a treatment group. At the beginning of the survey, all users need to fill in the OD pair, departure time, and average travel time of their own frequent morning trips. Then, they will see the instruction for the lottery questions. After passing a test, they will ask to answer 6 lottery questions for the car and the other 6 questions for PT, the order of car and pt questions will be random. Respondents in the control group make choices between lotteries, which need to guess the travel time from Coventry to Birmingham University. Respondents in the Information group will receive additional historical travel time of the past 4 trips of the same OD pair. Respondents in the individualized group will need to guess the travel time of their own morning frequent trips, without additional information. After these lottery questions, all respondents will also be asked about their willingness to pay for risk and ambiguous travel time by car and bus, and their social demographics.

Randomization Method

We use the randomization function in Qualtris

Randomization Unit

Individual

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

1200 participants

Sample size: planned number of observations

1200 participants

Sample size (or number of clusters) by treatment arms

400 participants control, 400 participants information, 400 participants individualized trip

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Unit of measurement: the value of ambiguity attitudes at the individual level, using the belief for the probability of PT time interval 1 (b1_PT) as an example. For a type I error probability of alpha = 0.05, a power of 1- k = 0.95, and the standardized minimum detectable effect size is 0.3, N =253 per treatment group.

Supporting Documents and Materials

IRB