ABSTRACT
This paper examines travelers’ preference for airline livery design, which is an important component of sensory experience for tourists. Based on surveys of both U.S. and South Korean respondents, this study finds that airline livery preference depends on travel purpose (i.e. business vs. vacation). Designs preferred by the two travel segments differ in terms of color tone, hue, line, and shape. Business traveler-preferred designs signal punctual, efficient, classy, and safe service; while vacation traveler-preferred designs elicit feelings of fun, friendly, and exotic service. It is also shown that travelers prefer airline livery that is congruent with their situational self-image, depending on the travel occasion at hand. In fact, self-image congruency of design is approximately half of the variance in airline livery preference. Finally, results show that self-image congruency also increases perceived service quality.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Notes
1. Terms congruency, congruence, and congruity are used interchangeably in previous research (Back, Citation2005; Hanks et al., Citation2017; Hosany & Martin, Citation2012; Kleijnen et al., Citation2005).
2. Tailfin design was used because it has been identified as the most visible component of an airline’s brand (Thurlow & Aiello, Citation2007).
3. VIF values for the business model were, 1.41, 5.89, 3.88, 3.85, 2.85, 3.39, 5.34, 3.24, 2.82, 4.23 for the independent variables, in the order of the variables in . The VIF values for the vacation model were 1.36, 5.30, 3.56, 3.41, 2.50, 3.41, 5.36, 3.04, 2.78, 4.13, respectively.
4. Examples of original wording from Sirgy et. al. (1997) were as follows. “ (a) This outfit is consistent with how I see myself at work. (b) This outfit reflects who I am at work. (c) People similar to me wear outfits like this at work.”
5. Indirect effects in the PROCESS macro are measured through the bootstrapping method, which does not assume normality of error terms. Therefore, indirect effects do not have a t-statistic like the direct effects. Also, when the 95% confidence interval does not include a zero, the effect is significant at α = .05.