Griffiths, R E & C Reschovsky: Geographically

Report
Geographically-Focused
Household Travel Surveys
in the
Metro Washington Region
(2010 – 2012)
Robert E. Griffiths & Clara Reschovsky
Metropolitan Washington Council of Governments
TRB Travel Survey Methods Committee Meeting
January16, 2013
Project Background
• Follow-on to 2007-2008 TPB Regional Household Travel
Survey that was primarily conducted for the
development of the new travel demand model
• Household Travel Survey data collection in specific
geographic sub-areas of the region (Case Studies)
• Addresses a need expressed by local planners
• Will provide some current small area community-level
socio-economic data that are no longer available from
the Decennial Census
1
Project Objectives
• Analyze daily travel behavior in communities with
different densities, physical characteristics and
transportation options - “Regional Activity Centers”
• Assist local planners with current local land use and
transportation planning efforts
• Build a household travel survey database that can
measure changes in local community travel behavior over
a period of time (Before and After comparisons)
2
Arlington County Spring 2010
Supplemental Survey Effort
Three areas in Arlington County, Virginia were surveyed:
• The Jefferson Davis Highway/Crystal City/Pentagon City Area
• The Village of Shirlington
• The Columbia Pike Corridor
Collaborative effort between Arlington County and COG/TPB
staff to supplement the 2007-2008 TPB Household Travel Survey
with additional household travel data collected in 2010.
Interested in knowing more about how new higher density
residential and commercial development was affecting daily
travel behavior in these areas.
3
Arlington County Subareas
Spring 2010
Jeff Davis Hwy/Crystal City/Pentagon City
Land Area
= 0.7 sq mi
Households
= 9,600
Population
= 15,300
Pop Density
= 22,300 persons/sq mi
Shirlington Area
Land Area
Households
Population
Pop Density
= 0.6 sq mi.
= 4,200
= 7,200
= 12,900 persons/sq mi
Columbia Pike Corridor
Land Area
= 2.5 sq mi
Households
= 15,000
Population
= 35,200
Pop Density
= 14,100
4
Now Have Collected Survey
Data in 17 Focused Areas
Dulles
Friendship
Heights
New York Ave
Falls Church
Beauregard
Corridor
National
Harbor
St Charles
5
Focused Geo-Areas
•
•
•
•
Area Size: 1 to 20 Square Miles (Most < 8 Sq mi)
Number of HHs: 12,500 to 25,000
Sampling Rate: 1 in 6 to 1 in 3
Sampling Method: Address-Based Sample with
Systematic Selection
• Target Completes: About 400 Completed Households
in each Geo-Area
• Assumed Response Rate: 10%
• Survey Period: Spring and Fall (2 -3 Months)
6
Survey Methodology Basics
Two-Stage Survey
(1) Recruitment Stage
– Advance Letter
– Request Participation
– Obtain Basic Information about Household
(2) Travel Data Retrieval Stage
– Travel Diary for 24-hr Weekday
– Retrieve Data on Daily Travel
for each Household Member
7
Recruitment Stage
• Address Sample is Phone-Matched
– 35 to 65% of Addresses Phone Matched
• Initial Advance Letter Mailing in English and Spanish w/$1 Bill
– Offer $25 Gift Card Incentive for Survey Completion
– Short 10-Question Questionnaire
• Roster Household Members & Vehicles
• Get HH Telephone Number & E-mail Address (Optional)
• Business Reply Mail-Back Envelope
• Reminder Post Card, Additional Recruitment package mailing
and 2nd Reminder Post Card
– Web Response for HH Questionnaire after 1st Reminder PC
8
Recruitment Stage: Results
• Postal Non-Deliverables ranged from 3% to 10% (5.5% avg)
• Mail-Back of HH Questionnaire by households receiving the
mailed recruitment materials ranged from 11 to 21% (15% avg)
• Web completion of HH Questionnaire < 1%
• About 2/3rd of the households returning the HH Questionnaire
were recruited to participate in the survey (range 61 to 76%)
• Phone recruitment of households not returning the HH
Questionnaire, but with known telephone numbers resulted in
another 3% to 4% of the households receiving the original
mailing
• Overall Recruitment Response Rate averaged 14%
9
Retrieval Stage:
• Almost all travel dairy survey data was retrieved via a CATI survey
interview
• Proxy Interviews conducted for children and call-backs
after 3-days
• Less than 1% of the households mailed back their survey
interviews
• Required retrieval of travel day data from every household
member to be considered a “completed” household interview
• Retrieval Rates ranged between 62% and 78% (avg. 72%)
• The overall response rate ranged from 7.0% to 16.3% (avg. 10.7%)
10
Results: Use of Survey
• Results well received Board and local planning staff
• Supports planning focus on Regional Activity Centers
• Window on small areas not previously available, provides a
dimension scale not seen in “Regional” averages
• Obtain travel modes not routinely well represented in regionwide household travel surveys
• Even modelers now beginning to show interest in the GeoFocused Survey data
11
Commute Mode Share 2010/2011
Drive Alone
(SOV)
Carpool
(HOV)
Transit
Walk
Bike
Other
Core
Logan Circle
21%
4%
28%
33%
10.6%
2%
Crystal City
22%
4%
53%
19%
0.7%
2%
Inner
Largo
70%
11%
13%
3%
2.8%
--
Reston
70%
17%
8%
3%
0.7%
2%
Outer
In Neighborhoods Throughout the Region
Woodbridge
76%
13%
8%
1%
0.3%
2%
Frederick
78%
12%
4%
4%
1.5%
--
* Based on initial findings from the 2010/2011 Geographically
Focused Household Surveys reported to the TPB on May 16,
2012
12
Lessons Learned (so far)
• Important to maintain compatibility with earlier surveys
• Incentives
– Checks are a hassle for many recipients
– Gift cards do not seem to be much better
– They cannot be used everywhere
• Cannot be used online
• Cannot be used in many restaurants, independent establishments or
to buy certain items like phone cards
– Cannot track usage for ‘lost cards’
– Need policy of one replacement only
– Options? Email for online GC with physical GC option
• Web Recruitment
– Start Web Recruitment with 1st Mailing to Household
13
Thoughts on Methodological
Improvements
• Increase Duration of survey period
– Longer period allows for adjustment of sample &
survey procedures, more travel dates
• Reduce Respondent Burden
– Newer technology can lower burden or perceived
burden
• Doesn’t work for HHs without tech toys/access
• Lower perceived burden isn’t always lower
– Participants want control – I will call you with my
travel
14
Questions?
Contacts:
Robert E Griffiths
[email protected]
Clara Reschovsky
[email protected]
15

similar documents