InternetBased Stated Preference Survey

1

• Internet-Based Stated Preference Survey
• Studies in Traffic Information and Public
  Transport
• Åse Nossum, Research Economist
• Institute of Transport EconomicsOslo, Norway

2
Why An Internet Survey?

• Less expensive
• Same advantages as other computerized methods
  (customized design)
• Ready to analyse
• High access rate in Norway
• Good Internet skills in Norway

3
Two studies

• 1 Public Transport Study
• Passengers in public transport and their
  valuation of time savingTarget group People
  leaving in the capital region
• 2 Traffic Information Study
• Focusing on the use of and the need for
  information connected to the trip to work in
  morning rush hourTarget group people who
  commute by car to work in Oslo and experience
  congestion problems

4
Public Transport Study

• Sample chosen randomly from the population
  register
• From age 14 and above and address in the capital
  area
• Letter by post - personal password/user name
• A hard copy version for those without Internet
  access
• Reminder by mail

5
Traffic Information Study

• Nearly 1800 cards were distributed at three
  different locations in Oslo during morning rush
  hours
• Personal password/username
• A hard copy version for those without Internet
  access
• No possibility to send a reminder

6
Survey design

• Stated preference analysis for web hypothetical
  choice situations (possible to introduce new
  choice situations).
• Sawtooth software (easy-to-use templates, skip
  patterns, response verification, show answers to
  previous question).

7
Example of a choice sequence.
If these trips were identical in all other
respects, which one would you choose?
Car trip A
Car trip B
Cost 1.80
Cost 2.26
Travel time 33 min
Travel time 27 min
Information of type A, B and C
Information of type A
Probably A
Definitely A
Dont know
Probably B
Definitely B
8
(No Transcript)
9
Findings

• Response rate on Internet of 16-18
• High data quality?
• Representative sample?
• 20 min to complete the questionnaire
• Many completed answers
• Only 7 persons completed the paper copy in the
  Traffic Information Study

10
Responses per day on Internet
11
Responses on Internet during the day
12
Questionnaire suitable for an Internet survey

• High access to Internet in the target group
• Not to complicated questions
• Self-explanatory questions
• Questionnaire of limited length
• Attractive design of the questionnaire
• Subject that engage people

13
The great challenge

• Representative sample
• Target group
• Internet surveys in combinationswith home
  interviews (Tønsberg)