Do satisfying walking and cycling trips result in more future trips with active travel modes? An exploratory study

ABSTRACT

Previous studies have indicated that travel satisfaction - the experienced emotions during, and cognitive evaluation of, a trip - can be affected by travel mode choice and other trip characteristics. However, as satisfactory trips might improve a person's attitude towards the used mode, persons may be more likely to use that same mode for future trips of the same kind. Hence, a cyclical process between travel mode choice and travel satisfaction might occur. In this paper we begin to analyse this process—using a structural equation modelling approach on cross-sectional data—for people who engage in walking and cycling for leisure trips in the Belgian city of Ghent. The focus on walking and cycling reflects recent studies indicating that active travel is often associated with the highest levels of travel satisfaction. Results of this exploratory analysis offer tentative support for the idea of a cyclical process: the evaluation of walking and cycling trips positively affects the respondents' attitude towards the respective mode, which in turn has a positive effect on choosing that mode.

KEYWORDS:

• Active travel
• residential location
• structural equation modelling
• travel mode choice
• travel satisfaction
• travel-related attitudes

Acknowledgments

We would like to thank the four anonymous reviewers for their constructive comments and suggestions, making it possible to improve this article. This research has been funded by the Research Foundation - Flanders (FWO), grant 12F2516N.

Notes

¹ Although it is possible that the choice of a certain mode is affected by attitudes towards other modes (e.g., cycling being affected by attitudes towards walking), we will only analyse the effect of mode-specific attitudes on the choice of that mode.

² 11 respondents who did not indicate which travel mode they used, were also removed.

³ We have tried to limit possible confounding of trip satisfaction with the liking for the activity at the destination by first asking respondents to assess the leisure activity itself, by asking information about the trip (e.g., mode choice) prior to the STS and by italicising ‘during your trip’ in the question ‘Which emotions did you experience during your trip towards your most recent out-of-home leisure activity?’

⁴ Average scores on affect and cognitive evaluation are respectively 1.16 and 1.33 for car users, and 0.91 and 1.21 for public transport users (bus/tram users and train users combined).

⁵ Degree of travel liking (min. 1, max. 5): cycling: 4.35 (4.70 for cyclists and 4.24 for non-cyclists); walking: 3.99 (4.43 for pedestrians and 3.88 for non-pedestrians); car use: 3.82 (4.09 for car users and 3.53 for non-car users); train use: 3.44 (4.15 for train users and 3.41 for non-train users); bus/tram use: 2.78 (3.61 for bus/tram users and 2.72 for non-bus/tram users). Number of positive aspects linked to travel mode (min. 0, max. 12): cycling: 6.47 (8.21 for cyclists and 5.96 for non-cyclists); walking: 5.73 (6.71 for pedestrians and 5.51 for non-pedestrians); car use: 4.26 (4.87 for car users and 3.69 for non-car users); train use: 2.90 (4.49 for train users and 2.84 for non-train users); bus/tram use: 1.94 (2.99 for bus/tram users and 1.87 for non-bus/tram users).

¹⁰ The (rotated) factor loadings in the pattern matrix represent the degree of association between the variable and the factor.

^,11 The variable ideal neighbourhood: high density has the highest loading on the Pro walking factor (i.e., 0.14), but has been suppressed from the table due to a loading between −0.25 and 0.25.

⁶ This would probably be the case for public transport users, since these people have - on average - the lowest levels of travel satisfaction (as is the case in our sample (see Section 4.1.2), but also in most other studies analysing the effect of mode choice on travel satisfaction (see Section 2.1)). A result of this (possible) negative reinforcement process would be that attitudes towards public transport are lower than attitudes towards other modes, which is found in our data set (see Section 4.1.3). Lower levels of attitudes towards public transport compared to attitudes towards other modes have also been found in other studies (e.g., Kroesen, Handy, & Chorus, 2017; van Wee, Holwerda, & van Baren, 2002). This negative reinforcement process will potentially result in low levels of public transport use (especially when destinations choice is mostly free such as with leisure trips; in our sample only 9.5% of the respondents used public transport), and/or a considerable amount of people using public transport while having preferences for other travel modes. This forced use could be the result of destinations being outside walking/cycling distance or not having access to a car.

⁷ If the stability index falls between –1 and +1, the system is stable.

⁸ Analysing squared multiple correlations of nonrecursive models can be problematic as negative squared multiple correlations can be obtained (Hayduk, 2006). However, this is not the case in our models.

⁹ The relationship between travel-related attitudes and the residential location has to be treated with caution since most respondents already live in their neighbourhood for multiple years (56.1% of the respondents indicate that they already live more than five years in their current dwelling), whereby both the built environment and the travel-related attitudes might have changed in relation to the time they made their residential location choice.

Additional information

Funding

Research Foundation - Flanders (FWO) (12F2516N).