Determinants of Intra-Destination Mobility of Tourists

This study investigates factors influencing tourists’ choices between passive and active modes of transportation during short-distance trips at holiday destinations. Therefore, we plan to run a large-scale artefactual field experiment with a national representative sample of the German adult population, where respondents take part in a discrete choice experiment (DCE).
Respondents will be asked to choose the mode of transport (walking, bicycle, e-bike, or car) for a hypothetical short-distance trip at the holiday destination. All respondents read a hypothetical scenario in which they are spending a summer holiday in a village in the mountains with another adult. They plan to visit a nearby lake and must choose between different modes of transportation. The scenario differs along several attributes: weather, infrastructure (bike-lane), distance, necessary registration for bike rental, costs for renting an e-bike, flexibility of return of rental bikes, parking costs, availability of parking, and possibility and distance of alternative (scenic) routes.
Respondents are presented with 6 different random scenarios and have to choose the mode of transport.

Respondents will conduct the experiment via an online platform.
Respondents will see six hypothetical scenarios with (randomly) drawn attributes (weather, infrastructure (bike-lane), distance, necessary registration for bike rental, costs for renting an e-bike, flexibility of return of rental bikes, parking costs, availability of parking, and possibility and distance of alternative (scenic) routes) and will have to choose the mode of transport (walking, bicycle, e-bike, or car). Here are the different attributes and the probability that they are drawn:
• Weather: Sunny (33%), Cloudy (33%), Light Rain (33%)
• Infrastructure (bike-lane): Separate bike lane (33%), Bike lane right next to main road (33%), Riding in mixed traffic (33%),
• Distance: 1 km (20%), 3 km (20%), 5 km (20%), 12 km (20%), 18 km (20%)
• Necessary registration for bike rental: No registration/app required (33%), Registration at reception (fill out form, show ID) (33%), Registration via app required (download app and register) (33%)
• Costs for renting one e-bike: 0€ (25%), 6€ (25%), 12€ (25%), 24€ (25%),
• Flexibility of return of rental bikes: Return possible only at the accommodation (50%), Return also possible at the lake (50%)
• Parking costs: 0€ (33%), 6€ (33%), 12€ (33%)
• Availability of parking: Enough parking (33%), Limited parking (33%), Few parking spaces (33%)
• Distance of alternative (scenic) routes: +1 km (33%), +3 km (33%), +5 km (33%)

Single Profile Discrete Choice Experiment

[Description wording:]

You are spending your summer vacation in a village in the mountains together with one adult companion. You arrived in your own car and stay in a bed-and-breakfast accommodation. In the afternoon, you would like to go to a nearby lake, spend some time there, and afterwards have dinner at the restaurant by the lake.

For the trip there and back, different modes of transportation are available to you. You can either use your own car, rent a bicycle or an e-bike, or walk. Regular bicycles are offered free of charge, while e-bikes can be rented for an additional rental fee.

At the swimming lake, parking spaces for cars as well as covered and secure parking facilities for bicycles are available.

If you rent a bicycle, you can ride it to the lake, and after dinner, you must either bring it back to your accommodation or leave it at the lake. If you leave it there, a free shuttle service will take you back to your accommodation. Whether you have to bring the bicycle back to the accommodation or can leave it at the lake varies from one decision to another.

The route to the lake is mostly flat. On foot or by bicycle, you can either take the direct route along the main road or an alternative route that runs along small roads and bicycle paths without car traffic through meadows and forests.

[Treatment Variation]
As explained above, a computer randomly draws attributes.

2026-03-19

2026-03-26

Choice of active mode of transport (walking, bicycle, e-bike) versus car.

We test 9 main hypotheses:
H1: Better weather increases the choice of active modes of transport
H2: Lower distance increases the choice of active modes of transport
H3: Safer bike lanes increase the choice of active modes of transport
H4: Low registration effort increases the choice of active modes of transport
H5: Higher parking costs for cars increase the choice of active modes of transport
H6: A lower number of parking slots increases the choice of active modes of transport
H7: Lower distance for the alternative route increases the choice of active modes of transport
H8: Lower rental cost increases the choice of active modes of transport
H9: Flexible bikes return option increases the choice of active modes of transport

To estimate the average marginal component effect of each attribute on the travel mode, we use the following regression model:

yij = α0 + α1∙weather + α2∙ distance + α3∙bikelane+ α4∙registration+ α5∙cost_parking+ α6∙slots_parking + α7∙detour_distance + α8∙costs_rental + α9∙flex_return + εijk

We define a dummy variable that is one if respondents choose an active mode of transport (walking, bicycle, e-bike) and zero if they choose the car.

Choice of bike (e-bike & bike) versus walking.
H2a: Lower distance increases the choice of walking

Choice of e-bike versus bike
H2b: Lower distance increases the choice of taking an e-bike
H7b: Lower distance for alternative route increases the choice of taking an e-bike
H8b: Lower rental cost increases the choice of taking an e-bike

To explore heterogeneity in preferences, we plan to conduct subgroup analyses focusing on respondents’ physical activity levels, cycling habits, and orientation toward active holidays.

We plan to sample a minimum of 2,500 adults aged between 18 and 69 years.

The experiment is conducted in cooperation with a survey institute. The recruitment and polling are managed by the survey institute. We collect the data via an online server (Heroku). Our participants take part in the experiment autonomously on their own digital devices. Randomization is carried out at the individual level, using a computer.

Our experiment is structured as follows:
Respondents will be seeing six consecutive hypothetical scenarios and must choose the mode of transport (walking, bicycle, e-bike, or car).

Randomization is carried out by a computer.

at the individual level

No

2,500

2,500 individuals (adults aged between 18 and 69 years)

2,500

None