The Swiss National Stated Preference Study on Transport 2015 Claude Weis (TransOptima GmbH) Matthias Kowald (RheinMain University of Applied Science) Kay Axhausen (ETH Zurich) Milos Balac (ETH Zurich) Nicole Mathys (ARE) May 2017
The Swiss Microcensus on Mobility and Transport Swiss national travel survey (MCMT); Conducted every five years (~60´000 participants); Sampling frame: Swiss population (aged 6 and above); Daily mobility (one day CATI diary survey); Stage based with route-recording during the interview (PMT & PT); Since 2010: follow-up stated preference (SP) paper and pencil questionnaire on mode and route choice.
Stated preference survey Sampling frame: MCMT respondents (aged 18 and above); Recruitment during MCMT interviews (~4´000 participants); Represents MCMT in trip characteristics (mode, purpose and length) and respondents´ spatial and socio-demographic properties; Aims to provide basis for transport policy decisions and updates of transport models by: surveying data on mode choice (HPM; PMT; PT) and route choice (PMT; PT); synchronizing two important data sources (MCMT & SP); employing familiar and realistic situations for respondents. vielleicht noch konkreter sagen, dass die RP-Daten für die Konstruktion der SP verwendet werden und dass bei der Auswertungen beide Datensätze verwendet werden können
Experimental design: Mode Choice Alternative Attribute Variation of empirical value PMT Travel time -20% / -10% / +30% Parking search time -25% / +25% / +50% Fuel cost -10% / +20% / +30% Road toll 50% / 100% / 200% of -.06 CHF/km Parking costs Delay risk 10% / 20% / 30% Delay time 50% / 100% / 150% of 0.1 x travel time (max. 30 min) PT In-vehicle time -30% / -10% / +30% Access / egress time -50% / +/-0% / +50% Travel costs No. of transfers -1 / +/-0 / +1 Headway -1 / +/-0 / +1 level* Capacity utilization -1 / +/-0 / +1 level** * Headway levels: 5, 7, 10, 15, 20, 30, 60, 90, 120 min ** Level of capacity utilization: low, median, high, overload Zu viele Ausprägungen, daher Aufteilung in zwei Attributsätze
Experimental design: Route Choice Mode Attribute Variation of empirical value PMT Travel time -30% / -10% / +30% Travel costs -10% / +20% / +30% Road toll 50% / 100% / 200% of -.06 CHF/km PT Main transportation mean -1 / +/-0 / +1 level In-vehicle time Access / egress time -50% / +/-0% / +50% No. of transfers -1 / +/-0 / +1 Transfer time Headway -1 / +/-0 / +1 Stufe Capacity utilization Ebenfalls Aufteilung der Routenwahl ÖV in zwei Attributsätze Bei 5 & 6 sagen, dass die Prozentzuschläge sich jeweils auf die aus dem MATSim-Modell abgeleiteten Attribute beziehen.
SP-questionnaire: Capacity utilization PT low medium intercity intercity local local high overload intercity intercity local local
Process of data collection Week 1 2 3 4 5 Procedure Day Mon – Sun Mon Tue Wed Thur Fri Mon – Fri Recruitment Data processing Monday Aggregate stages to trips Tuesday Choosing trips for SP experiment Wednesday Routing of trips (PMT & PT) Generate questionnaires (experimental design) Thursday Print & send questionnaires Friday Send reminders 14 Befragungswochen in der Hauptstudie evtl. Folie löschen, falls Zeit knapp
Response time and rate 14 Befragungswochen in der Hauptstudie
Item non-response in SP-experiments 14 Befragungswochen in der Hauptstudie
Representativeness: Age, Region, Distance, Purpose etwas viel Information für eine einzige Folie
Preliminary model results: Parameter ratios Variable Ratio vs. travel time [-] Willingness-to-pay [CHF/h] PMT Travel time - 15.2 Search time 1.92 39.4 Delay 1.68 25.6 PT 15.9 Access / egress time 1.34 21.2 Transfer time 0.73 11.6 No. of transfers 7.53 2.0 Headway 0.38 6.0 1.46 23.1
Model results: Values of travel time savings (VTTS) evtl noch das confidence interval anzeigen
Conclusions Connecting MCMT and SP experiments allows to: synchronize data sources; survey representative information; employ familiar and realistic situations; use synergy effects. Data are a reliable source for transport policy decisions and updates of transport models. Aggregation of MCMT stages to SP trips is challenging for some participants. Preliminary analysis results indicate only minor changes in parameters between SP 2010 and 2015. Deeper analysis is needed: purpose specific models; non linear models; complex models (MMNL; LC Models). die hier aufgeführten conclusions sind prozedural, technisch. Eine Folien zu den inhaltlichen Conclusions - Resultate, wie sie zu anderen Studien stehen und was es für die Verkehrspolitik bedeutet - wäre ein grosser Mehrwert "Aggregation of MCMT stages to SP trips is challenging for some participants." - müsste man vielleicht noch etwas ausführen