Travel-related feelings: review, theoretical framework, and numerical experiments

ABSTRACT

Limited previous research shows that travel by different modes evokes feelings. Also after-effects due to stress have been observed. Such travel-related feelings are important to consider in transport planning because of their possible consequences for travelers’ emotional well-being. A theoretical framework is proposed that makes quantitative predictions of the impacts of transient feelings (emotional responses) on enduring feelings (current mood) with consequences during and after travel. Positive and negative emotional responses are posited to be evoked by transient critical incidents (e.g. disruptions) and non-transient factors (e.g. noise) during travel. Numerical experiments illustrate the quantitative predictions on current mood during and after travel for both types of evoking factors.

KEYWORDS:

• Travel
• satisfaction
• feeling
• emotional response
• current mood
• emotional well-being
• theoretical framework
• numerical experiment

Acknowledgements

Preparation of this paper was made possible by grant 2014-05335 from the Swedish Governmental Agency for Innovation Systems awarded to the Service Research Center (CTF) and Samot VINN Excellence Center at Karlstad University, Sweden. I am grateful to my collaborators Dick Ettema, Margareta Friman, and Lars E. Olsson for valuable comments and to Michael Ståhl for assisting me with the numerical experiments.

Notes

1. Several different methods have been used to measure core affects including self-reports, startle responses, peripheral physiology, face expressions assessed by automated picture recognition systems or as electrical muscle potentials and brain measurements. A dimensional description of core affects is supported by these methods, although not all converge on the two orthogonal pleasure-displeasure and activation-deactivation dimensions (Mauss and Robinson 2009).

2. Västfjäll, Gärling, and Kleiner (2001) showed how a unidimensional dimension of positive vs. negative evaluations may be related to the pleasure-displeasure and activation-deactivation dimensions. See also the discussion in Kuppens et al. (2013) of different conceptualizations of the relation between pleasure-displeasure and activation-deactivation.

3. In empirical research (e.g. Eid and Diener 2004) mood has been shown to not vary substantially over time. This is accounted for in Equation (1) by assuming that mood does not change proportionally to the evaluation and by assuming that not all evaluations evoke emotional responses.

4. Prospect theory is currently used in many models of travel behavior (Li and Hensher 2011; Van de Kaa 2010). Note however that prospect theory accounts for evaluations of quantitative information but not for emotional responses or influences on mood. Equation (1) is thus a necessary complement.

5. This is referred to as the peak-end rule which has been demonstrated in many laboratory experiments (e.g. Schreiber and Kahneman 2000; for review, see Fredrickson 2000) although not as clearly in field studies (Kemp, Burt, and Furneaux 2008; Miron-Shatz 2009). A travel behavior example is Suzuki et al. (2014) who showed that satisfaction with commute trips to and from work is proportional to the average of duration-weighted satisfaction with the different stages of the trips (e.g. walking to the bus stop, riding the bus, walking to the work place). A caveat is that the ratings of the stages were made retrospectively at the same time as the retrospective ratings of the whole trip were made.

6. For simplicity the effects on current mood are analyzed at the last time segment of the journey.

7. The values of the parameters a _B, a _G, and b are selected to be approximately consistent with the findings in empirical studies (e.g. Carter and McBride 2013; Tversky and Kahneman 1992).