Data Collection MethodsIn Transportation Planning
Many data collection methods
• Customer satisfaction: http://www.dot.ca.gov/hq/maint/external_survey/2005_survey/index.htm
• Public involvement : http://www.101inmotion.com/
Data Collection Methods• Traffic monitoring: Loop
detector data (presence, speed, vehicle class)
• Used to provide historical records of use and real time information to motorists: http://www.dot.ca.gov/traffic/
• Traffic cameras: http://video.dot.ca.gov/
Travel SurveyQueen of Data Collection Methods
Design of a Survey• It is a process with many feedback
loops• Many choices depend on the reason
we collect data– Data as the base inventory– Data for models and simulations– Data for public involvement– Data for monitoring change (before-after)
• It is heavily influenced by time and money budgets
• Heavily quantitative but qualitative emerges as potentially powerful
• It reflects the designers model of society or explicit/implicit models of funding agencies
Survey ProcessRichardson-Ampt-Meyburg
Survey Methods for Transport PlanningEucalyptus Press
Human InteractionEmerging Research Direction in Travel
Behavior
Male, 40 years, full time worker, driver
Female, 36 years, home duties full time, driver
Male, 6 years, primary school
Male, 5 years, primary school
A family’s day in South Perth
Figure by Pribyl
Male, 47 years, full time worker, driver
Female, 42 years, home duties full time, driver
Female, 8 years, primary school
Male, 5 years, primary school
Another family’s day in
South Perth
Figure by Pribyl
10 Steps to Successful Surveys (Salant & Dillman)
• Understand and avoid the four kinds of error• Be specific about what information you seek
and why• Choose the survey method that will work
best• Decide whether and how to sample• Write good questions• Design and test the questionnaire• Assemble the necessary resources• Code, computerize and analyze the data• Present your results• Maintain perspective• Reports and presentations
Coverage Error• Occurs when the list (sampling
frame) from which the sample is drawn does not include all segments of the population that we want study/model
• The list used is not accurate and up-to-date
Sampling Error• Occurs because the survey
does not cover the entire population – it is not a census
• Always happens to some extent• Reduces by increasing the
sample size
Measurement Error• Occurs when a respondent’s
answer to a question is inaccurate, imprecise or in any other way unusable
• These errors come from— the survey method, the questionnaire, the interviewer, and the respondent
• Relatively easier for the researcher to control using statistical methods
Non-Response Error• Occurs when a significant
number of people in the sample do not respond to the questionnaire and are different from those that do in some way (this is also named selectivity bias in models)
• This is a main reason to maximize the response rate
Main Types of Surveys• Mail surveys• Telephone surveys• Face-to-face surveys• Internet surveys (immobile and
mobile)
When to Use Mail Surveys• You have a reliable address list• The people in the population
will respond accurately and completely by mail
• An immediate turnaround is not required
• Resources are relatively scarce but you have considerable amount of time
When to Use Telephone Surveys• The members of the population
are very likely to have telephones
• The questions are relatively straightforward
• Experienced help is available• A quick turnaround is
important
When to Face-to-Face Surveys• There is no list of the
population• You need to collect information
from people who are not likely to respond willingly or accurately by mail or telephone
• The questionnaire is complex• Well-funded projects• Experienced interviewers are
available
Maximize the Reward for Responding
• Show the respondent positive regard
• Give verbal appreciation• Use a “consulting” approach• Support the respondents
values• Offer tangible rewards• Make the questionnaire
interesting
Minimize the Cost of Responding
• Make the task appear brief• Reduce the physical and
mental effort required• Eliminate chances of
embarrassment• Eliminate any implication of
subordination• Eliminate any direct monetary
cost
Establish Trust With the Respondent
• Provide a token of appreciation in advance
• Identify with a known organization that has legitimacy
• Build on other exchange relationships
Schedule for the next few weeks
• Today (10/16/06)– Section survey type
• Next week (10/23/06)– Data collection methods part 2… components
of surveys. – Section broken into groups – Start writing questions
• 10/30/06– Issues with components, continue writing
questions
• 10/31/06 – Check for redundancy in survey (lecture)
• 11/3/06 (Friday) – Component due to me
Questions for reading about survey type• What is the Daily Reconstruction
Method (or Time Use Diary)?• What are the advantages of this
survey type?• What are the main categories or
sections (components) of the survey?
• Start thinking about how you would implement this technique in transportation (challenge: DRM in transportation setting, time use with attitudes and perceptions).
Write a 1 page MAX summary about your survey method with these questions in mind.
EMAIL TO ME BY THIS FRIDAY!!!!!!!