National Travel Survey Great Britain Presentation to Eurostat
National Travel Survey Great Britain Presentation to Eurostat 17 June 2013 Lyndsey Melbourne Department for Transport, UK
Agenda Background to the ‘National Travel Survey (NTS)’ in Great Britain Sample design, training & fieldwork Data collection Data linking Weighting Publication / Users of NTS data Future of the NTS (inc. GPS pilot) Questions
Background to the NTS Large household survey First survey in 1965. Continuous since 1988 National Centre for Social Research (Nat. Cen) became the contractor in 2002 Sample size tripled in 2002 to allow single year analysis Overall response rate is around 60% each year § achieved: n = 8 k households, 19 k people Weighted for non response since 1995 All household members, including children
Continuous survey The UK Government commissioned the first NTS in 1965/1966, and it was repeated on an ad-hoc basis in 1972/1973, 1975/1976, 1978/1979, and 1985/1986. In July 1988, the NTS became a continuous survey (i. e. fieldwork was conducted on a monthly basis). Advantages: § Up-to-date evidence available for policy development and monitoring travel trends § Lower set-up costs § Higher quality data – interviewer expertise retained, higher response rates
Why is the NTS important? Measures long-term trends in personal travel behaviour within Great Britain
Why is the NTS important? The only national data source available to answer key questions on: How? Walk, cycle, bus, car, train… Why? Commute, travel to school, shopping… Who? Children, elderly, low/high incomes, with/without a car, Very detailed dataset, allowing analysis of travel patterns by different groups
How is data collected? Placement and pick-up interview (CAPI) 7 day travel diary
Sample design Stratified two-stage random probability sample of 15, 048 private households in Great Britain, drawn from the Postcode Address File (PAF) § Stratified by region, car ownership and population density 684 Primary sampling units (PSUs) sampling 22 addresses each 57 PSUs assigned per month London PSUs are over-sampled as response rates tend to be much lower
Training and liaison Two day briefing for all new interviewers Annual ‘refresher’ briefings for all interviewers held in November/December each year in each region of the country Monthly liaison meetings between Df. T and Nat. Cen project managers. Discuss monthly response rates, quality indicators, any issues, etc
Fieldwork Advance letters sent by interviewers at start of month. Includes a book of 1 st class stamps (incentive). Interviewers make contact with households by personal visit. Arrange time/date for face-to-face placement interview. Travel Week Allocation Card used to ensure even spread of diary start dates. Mid-week check (half way through travel week). Pick-up interview (within 6 days of end of Travel week). £ 5 gift card for ‘fully responding’ households (incentive).
Interview Data: household, individual, vehicle Household data e. g: Individual data e. g: household structure, age, gender, vehicle availability, employment, education, access to key services, driving licence, bicycle own, satisfaction with transport, ticket types/passes, income, shopping frequency of travel by mode, working from home, Vehicle data e. g: Registration number, age, travel difficulties, engine size, fuel type, make & road accidents model, company/private, mileage, parking, Sat. Nav
Diary Data: Travel information For each trip: purpose start/end time origin/destination mode (for each stage) time distance number in party cost Day 7: short walks (less than 1 mile)
Long Distance Journeys (LDJs) LDJs – trips of 50 miles and over within GB Collected in the diary and during the interview Collected over longer time period (recall) – 3 weeks 1992 to 2005 – 1 week from 2006
Mileage card Vehicle mileage in travel week recorded for each vehicle Start and end milometer readings
Data structure
Why is the survey so detailed? There is a need for information at the following levels to be able to answer particular questions: Household Individual Vehicle Trip Stage How many trips are made per individual per year in households with no car, 1 car and 2+ cars? What is the annual car mileage in the lowest and highest income households?
Data linking - Vehicles Vehicle registration number matched to DVLA database to obtain: – Month & year registration, e. g. April 2010 (used to derive age) – Make and model, e. g. VOLKSWAGEN POLO 1. 4 CL – Length, e. g. 3715 mm – Engine Size, e. g. 1390 cc – CO 2, e. g. 205 g – Propulsion Type, e. g. petrol High quality data & reduces respondent burden In future we may use the MOT test database to obtain annual mileage – Odometer readings are recorded annually as part of test – Only for vehicles over 3 yrs old
Data linking - Address Household address is matched to the Postcode Directory to obtain: – – Urban/rural classification Settlement size Index of Multiple Deprivation Socio-demographic classifications In future we will link addresses to GIS and travel time datasets to obtain: – Walk time to nearest bus stop, train station, etc – Journey time (by mode of transport) to key services, e. g. hospital, supermarket, school, etc
Data linking - Gazetteer Diary data: origin and destination of each trip is coded to a geographic location using a Gazetteer. LDJ information from interview: origin and destination. Work place location.
Weighting methodology developed in 2005 and applied to NTS data back to 1995 Weights adjust for: § non-response bias § trip recording drop-off § Long Distance Journey interview data
NTS Technical Report https: //www. gov. uk/government/uploads/system/uploads/attachment_data/file/36548/nts 2011 -technical. pdf
NTS Statistical Release www. gov. uk Also publish NTS factsheets on a range of subjects including Commuting & business travel, Walking and Travel to school
Travel behaviour in 2011 (compared with 1972/73) 958 trips person per year +0. 2% 364 hours spent travelling pppy +3% 6, 826 miles distance travelled pppy +53%
Recent uses within Df. T High Speed 2 – NTS data underpins Df. T modelling tools which contributed to the economic business case. NTS results were included in support materials, e. g. leaflet showing charts of trend in long-distance rail travel and socio-economic profile of rail travellers. Analysis in support of the Government’s 2010 Spending Review and the 2012 Autumn Statement, notably in highlighting the ‘distributional’ impact of proposals on different social groups Monitoring a broad range of transport policy indicators, including walking and cycling rates, private and company car mileage, concessionary bus pass take-up & usage, driver licence holding, travel to school, etc.
Other users of the NTS Central Government: Department for … – Environment, Food & Rural Affairs – Energy & Climate change – Health – Education The Treasury & HMRC Scottish Government & Welsh Assembly Competition Commission Transport for London Transport Select Committee & Parliamentary Advisory Council for Transport Safety Various county councils & local authorities Cyclist Touring Club Sustrans RAC & AA & Motorcycle Industry Association Car companies Campaign for Better Transport Passenger Transport Executives Passenger Focus Energy Saving Trust Age UK Commission for Rural Communities Developers/planners Consultancies/Academics (often working for local authorities or other government departments)
On the Move Report, RAC Foundation (3 December 2012 ) “The analysis reported here has drawn mainly on National Travel Survey (NTS) data, which, with its seven-day travel diary dataset, provides a unique source of information that is well suited to such in-depth analysis; it would not have been possible to uncover the insights presented in this report using any other existing dataset. . ” http: //www. racfoundation. org/research/mobility/on-the-move-main-research-page
Future of the NTS Public consultation on the future design of the NTS: June-Sept 2011 The proposals covered four areas: § Moving to a GPS methodology § Reducing the length of the questionnaire § Geographic coverage § Amending the methodology of the survey to make cost savings. GPS data collection pilot Drivers were: § Reducing respondent burden & costs § Improving data quality § Modernising methods
GPS data collection pilot Mobitest GPS logger with accelerometer; Collected data from nearly 900 respondents aged 12+, in Feb & March 2011 (sub-sample of main survey); Slighter lower response rate (52% for GPS pilot, 59% completed diary). GPS data processed by Eindhoven University of Technology (TU/e). Analysis of results - comparing diary data for same period to GPS data.
GPS data collection pilot GPS version of the NTS does not produce similar results to diary collection. § Fewer trips and stages in the GPS data. § Time taken and average trip length was longer. § GPS data had more trips to/from home and fewer trips to/from work than diary; 25% of trips were missing either a 'to' or 'from' purpose code. § Suspect GPS processing identified 'tours' rather than trips. Data processing market is not yet ready for use in an NTS context where mode and purpose are required.
GPS data collection pilot 3 published reports on the fieldwork, data processing and summary analysis of the 2011 NTS GPS pilot: National Travel Survey 2011 GPS Pilot Field Report (Nat. Cen) Processing of National Travel Survey GPS Pilot Data, a technical report (TU/e) National Travel Survey 2011 GPS pilot: a summary analysis (Df. T) https: //www. gov. uk/transport-statistics-notes-and-guidance -national-travel-survey
Future of the NTS 7 -day travel diary Contract with Nat. Cen renewed for fieldwork 2013 -2017 Reduced costs (34% saving) and respondent burden: § Shorter interview (removed questions) § Items removed from diary, e. g. tolls/congestion charge, children playing in street on day 7, taxi costs § England residents only sample, no data collection in Scotland Wales from 2013
Thank you for listening Any questions? Contact details Email: national. travelsurvey@dft. gsi. gov. uk Website: https: //www. gov. uk/government/organisations/department-fortransport/series/national-travel-survey-statistics Phone: 020 7944 3097
- Slides: 32