Tourism Management 80 (2020) 104108

Contents lists available at ScienceDirect

Tourism Management
journal homepage: http://www.elsevier.com/locate/tourman

Traveler preferences from online reviews: Role of travel goals, class

and culture
Swagato Chatterjee a, *, Prasenjit Mandal b
a
Indian Institute of Technology Kharagpur India, IIT Kharagpur, Kharagpur, West Bengal, PIN-721302, India
b
Indian Institute of Management Calcutta India, IIM Calcutta, Diamond Harbour Rd, Joka, Kolkata, West Bengal, 700104, India

A R T I C L E I N F O A B S T R A C T

Keywords: Along with being critical performance indicators as well as a rich source of information for potential customers,
Overall ratings the overall rating and recommendation behavior also work as a crucial information source for managers of
Recommendation airlines for customer segmentation, service design, and post-purchase evaluation management. In this study, we
Culture
explore how customer preferences, behavior and post-purchase evaluations differ for travelers depending on
Travel goals
Construal level
their respective travel goals, travel class, and the culture of their native country. Based on the frameworks of the
construal level theory and expectation disconfirmation theory, we provide a psychological explanation of why
specific service attributes, categorized in process-outcome and tangible—intangible dimensions, get higher
importance than others for a particular set of customers. We analyzed consumer review data of 28,341 reviews
for 345 airlines by customers from 132 countries. The results yielded critical theoretical contributions to extant
literature on airline evaluations and recommendations, construal levels, airline customer segmentation, etc.
Moreover, the results also helped managers of an airline to package design, segment customers, enhance targeted
communication, and customer relationship management overall. The limitations and future scope of the study
have also been discussed.

1. Introduction managing customer satisfaction and loyalty becomes more challenging

for marketers (Ostrowski, O’Brien, & Gordon, 1993). Moreover, man­
Formulation of a successful marketing strategy requires a proper aging airline service within the ambit of a multi-cultural platform would
understanding of customer preferences and perceptions about various require the marketers to know and understand socio-cultural prefer­
aspects of product and/or service. Such an understanding helps the ences and nuances virtually of every customer (Reisinger & Crotts,
marketers in handling positive and negative feedback, and also in 2010). Same is the case with the preferences of various segments of
formulating optimal product/service design that would essentially help customers; again, due to the introduction of low-cost airlines today,
them in the long run (Lin & Vlachos, 2018). Moreover, knowledge of customers from different socio-economic backgrounds, often with
such varied preferences of the customers can help the service providers different travel goals fly across states and countries (Hwang & Hyun,
create optimal segmentation strategy so that they can target the 2017; Hwang & Lyu, 2018). Managing all these in amalgamation under
customer segments properly using the right positioning. In fact, this a single study to understand the varied preferences of the consumers is
knowledge of varied preference of the consumers is crucial especially in very difficult for both practitioners and researchers. The current study
the context of airlines, more so today than ever before; introduction of tries to do the same. The aim of the study is to explore the varied pref­
low-cost carriers in the recent past have brought about significant erences of the airline customers based on their travel goals, travelling
change of customer behavior, increased the competition to such an class and cultural background. We used construal level theory and ex­
extent that customer satisfaction and loyalty have become the only way pectancy disconfirmation theory to propose the theoretical foundation
of survival (Akamavi, Mohamed, Pellmann, & Xu, 2015; Park, Rob­ of the above study.
ertson, & Wu, 2004). On the other hand, airline service being a complex Extant literature has majorly used consumer survey responses as the
combination of multiple service aspects such as check-in, food and source of data while studying consumer preferences which can be used
beverage, in-flight service, comfort, safety, entertainment, etc., for customer segmentation. However, with growing importance of word-

* Corresponding author.
E-mail addresses: [email protected] (S. Chatterjee), [email protected] (P. Mandal).

https://doi.org/10.1016/j.tourman.2020.104108
Received 19 December 2019; Received in revised form 3 March 2020; Accepted 17 March 2020
Available online 31 March 2020
0261-5177/© 2020 Elsevier Ltd. All rights reserved.
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

of-mouth in consumer decision making, studies started focusing on Hernandez-Ortega, and Brethouwer (2013) studied the cross cultural
another vital source of such information that is available today is differences in online airline ticket purchasing behavior among con­
customer reviews and recommendations, which reflect the customer’s sumers from Spain and Nertherlands. All these studies were limited
perceptions in detail (Chatterjee, 2019). Customer reviews and recom­ either in terms of choice of method (survey), selection of countries
mendations help in customer’s net promoter score and customer loyalty (mostly western and eastern) and/or the number of different cultures
(Ho-Dac, Carson, & Moore, 2013), which are essentially the indicators of they could effectively include, which are very. Moreover, no psycho­
long-term profitability of an organization (Reichheld, 2003; Vlachos & logical mechanism driving the cross cultural differences were also
Lin, 2014). With the boom of the internet usage today, customer reviews studied in these papers. Our study tries to fill this gap by providing a
and word of mouth (WoM) are more common in online platforms, where cross-cultural study of customer evaluations and recommendation
customers express their views and opinions, rate services in multiple behavior using a large real dataset of user-generated content consisting
parameters and encourage or discourage others for purchasing a prod­ of customers from different parts of the world along with their cultural
ucts and/or service (Chatterjee, 2019; Park, Gu, Leung, & Konana, background. We used construal level theory and expectancy disconfir­
2014). Such massive growth in data availability makes an opportunity mation theory to provide the theoretical background of such
for marketers to create insights about customer preferences and per­ cross-cultural differences. Table 1 provides comparative positioning of
ceptions based on online user-generated content. Moreover, due to the the current study in extant literature of cross-cultural studies on airline
large number of fliers crisscrossing states and countries every day, col­ passengers.
lecting data about their preferences become a painful exercise, as it in­
volves culling data from multiple countries and thereby cultures and 2.2. Airline customer segmentation – purpose of travel and class of travel
socio-cultural background. Consumer reviews and recommendation
posted in online platforms can help in this aspect. Another stream of literature where this study contributes is the
In this study, we have used user-generated content from travelers segmentation of airline customers. In a highly competitive market as the
expressing their evaluations and recommendations about airlines. The airline industry, where low-cost airlines have significant market share
size of the data collected vis a vis the real nature of the data in itself, and high growth rate globally, knowing customer preferences while also
enabled us to explore differential preferences of customers based on segregating them correctly is essential (Doganis, 2006; Teichert, Shehu,
their culture, socio-economic backgrounds, and travel goals. Moreover, & von Wartburg, 2008). Traditionally, it was seen that business travelers
we also came to know the relative importance of various service aspects, have been less price sensitive; however, in recent times, even they seem
categorized in process-outcomes and tangible-intangible dimensions for to buy tickets based upon its price (Mason, 2002; Teichert et al., 2008).
some of the customer-segments mentioned above. Additionally, we have On the other hand, due to additional flexibility and low costs, the fre­
also been able to provide a theoretical explanation of the differential quency of travels by leisure travelers went up to a point where customers
preferences of the segments, which is essential for theory building, and took their own decisions of travel plans (Chang & Hung, 2013; Mason,
thereby contributing to travel and tourism literature. 2005). Therefore proper customer segmentation is still a challenge in
In the next part of the paper, we review extant literature to look for airline industry. Nevertheless, extant literature has not focused much on
gaps and suggest ways to fill them. Then we develop a theoretical model, how an airline segments are different in terms of their preferences. Huse
build hypotheses, followed by the discussions on methodology and and Evangelho (2007) focused on how business travelers have hetero­
analyze the results of our data. Next, based on the analysis and our geneity in themselves depending on their preferences and behavior.
findings, we discuss the theoretical and managerial implications, and Teichert et al. (2008) for instance, studied the differential preferences of
explore how they could help in decision-making. We conclude the paper business vs. leisure travelers and business class vs. economy class trav­
by highlighting the limitations of the current study and also mention its elers. Wen and Lai (2010) on the other hand, used choice data of cus­
future scopes. tomers along with latent class models to find the two segments and
differential preferences on service attributes of these segments. Marti­
2. Literature review nez-Garcia and Royo-Vela (2010) and Lu (2017) explored the segments
in low cost and full service airlines to find some specific traveler
This study is related to several streams of literature, filling thereby behavior of specific segments. Cho and Min (2018) focused on seg­
several gaps. In the next few sections within the literature review, we mentation of US based customers, but ignored cross-cultural segmen­
specifically focus on these two streams of literature while explaining the tation. What remains common among all these studies is that they have
research gaps, which the current study has attempted to fill. used survey data to find the differential preferences of service attributes.
One must note at this point that a survey has common method bias,
2.1. Cross-cultural study in airlines non-response bias along with several other limitations, which were
present in the studies mentioned above (Gideon, 2012). At our end, we
Role of culture within the service evaluation process has remained an have used actual consumer review data which have been accepted to be
important area of literature for years (Arnould & Thompson, 2005; more of a true representation of actual consumer experience and eval­
Mattila, 1999; Steenkamp, 2019). Erstwhile researchers have contrasted uation than primary survey data collected in the above studies, often
the preferences of ‘Western’ and Asian customers while choosing ser­ using imaginary situations or retrospective sampling (Chatterjee, 2019).
vices (Arnould & Thompson, 2005; Mattila, 1999; Steenkamp, 2019). The theoretical understanding of the reasons for the difference of pref­
However, in the context of airlines, cross-cultural studies have remained erence was also not present. Herein, along with nationality and culture,
limited. In a study on consumers from UK, Germany, Japan, Brazil and we focus on two other types of segmentation, based on travel goals and
Taiwan, Crotts and Erdmann (2000) found national culture has signifi­ based on travel class, and use actual user-generated data. We chose these
cant impact on consumer evaluations. Kim & Prideaux (2003) have because airline, often being the easiest gateway to international travel,
found quite a bit of difference amoung airline passengers in terms of carries passengers who are from different cultures. Moreover consumer
inflight and on-ground behavior based on their nation- Japan, Korea, preferences about various attributes is expected to be very different
China, and the US. Kim and Lee (2009) studied the cross cultural dif­ depending on how the attributes match the purpose of the travel and
ferences of airline passengers in their reactions to airline irregularity. how the attributes match the expectations based on price and quality
Reisinger and Crotts (2010) studied consumers from eight cultures and levels assured. Thus the choice of variables to segment the customers is
suggested, although cultures show the central tendency of the consumer interesting and relevant. Moreover, we focus on post-purchase satis­
behavior, cultural dimensions are bigger than national boundaries and faction and recommendation instead of purchase choices which is also
subcultures do exist within the boundary too. Ruiz-Mafe, Sanz-Blas, rare. Thus, our study fills the gap mentioned above in the literature of an

2
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

Table 1
Comparative positioning of this paper in extant literature on cross-cultural studies on airline passengers.
Author and Year Dependent variable Independent variables Method Cultures/Countries Psychological process used
Considered

Crotts and Evaluation and loyalty National Culture ANOVA UK, Germany, Japan, Not explored
Erdmann (2000) towards flights Brazil and Taiwan
Kim and Prideaux Airline passenger National Culture Qualitative Japan, Korea, China, and Not explored
(2003) behavior the US
Kim and Lee (2009) Response to airline National Culture ANOVA South Korea, Japan, China, Not explored
irregularity and America
Reisinger & Crotts, Hofstede cultural Nationality Exploratory Eight countries Not explored
2010 background
Ruiz-Mafe et al. Airline ticket online Perceived control, subjective norm SEM Spain and Netherlands Theory of planned behavior
(2013) purchase intention and attitude positively
Our study Airline evaluation Airline attributes, Cultural Regression 132 countries Construal level theory and Expectancy
dimensions disconfirmation theory

air traveler segmentation. pleasures of the journey. Therefore business travelers are less connected
with experiential satisfaction (Kashyap & Bojanic, 2000). As the satis­
3. Theoretical framework faction from a business travel is more associated with the business
success than the travel itself and as business success are often achieved
3.1. Construal level theory and customer evaluations in the long term; business travelers are expected to be in a higher psy­
chological distance with the travel experience, leading them thereby to a
Construal level theory differentiates attributes as central and pe­ higher construal level.
ripheral (Liberman, Trope, & Wakslak, 2007; Trope and Liberman On the other hand, leisure travelers’ major goal is to get pleasure
2010). While a central attribute refers to the gist of the activity and from the travel experience and travel service attributes (Kashyap &
answers the ‘what’ or ‘why’ questions, peripheral attribute refers to the Bojanic, 2000). Moreover, leisure travelers’ travel goals are more real, i.
‘how’ descriptions, and represents the process of the activity per se e. ‘then and there’ oriented. Leisure travelers therefore are expected to
(Higgins & Trope, 1990; Kruglanski, 1975). Moreover, the central ‘what’ be psychologically closer to the travel experience leading them to have a
or ‘why’ attributes are also related to the super-ordinate level of stimuli low construal level. It may be noted that customers, who are in a higher
identification (Liberman, Sagristano, & Trope, 2002), while the pe­ construal level, tend to be more generous while evaluating the service
ripheral ‘how’ attributes are associated with a subordinated level of (Huang et al., 2016); therefore, business travelers are expected to be
stimuli identification (Vallacher & Wegner, 1987). giving better overall ratings and higher recommendations. We posit
Extant literature has highlighted a positivity bias for customers in thereby:
high-construal. The positive aspects of experiences are more salient and
H1. Business travelers (high construal) gives (a) better overall rating
accessible for people who are in high construal group (Eyal, Liberman,
and (b) recommends more than leisure traveler (low construal)
Trope, & Walther, 2004; Herzog, Hansen, & W€ anke, 2007; Huang,
Burtch, Hong, & Polman, 2016). Extant researchers have found that part As discussed before, consumers in a lower construal level, tend to
experiences lead to more positive outcomes when they are described give more importance to the ‘how’ attributes or the process attributes,
with higher-construal levels (Williams & Bargh, 2008). The above re­ while consumers in a higher construal level, give higher importance to
sults were also replicated in the context of online customer reviews and the outcome attributes (Tatavarthy et al., 2019). We, thereby expect that
for temporal and spatial distances (Huang et al., 2016). Based on the business travelers, being in the higher construal level would naturally
discussion above, we conclude that a higher construal level does lead to provide higher importance to outcome attributes, and lesser significance
more favorable customer evaluations and actions. to process attributes, when compared to leisure travelers. Our next hy­
Extant literature has also shown that customers, who have a lower pothesis thereby is:
psychological and temporal distance to a purchase or consumptions
H2. Business travelers (high construal) gives (a) higher importance to
instance, would focus more on ‘vivid,’ tangible, immediate benefits. On
outcome attributes and (b) lesser importance to process attributes than
the other hand, customers, who have a higher psychological and tem­
leisure travelers (low construal).
poral distance to a purchase or consumptions instance, would focus on
distant, intangible, abstract attributes (Ding & Keh, 2017; Herna �nde­
z-Ortega, 2018). Researchers have also divided service attribute to 3.3. Time orientation of culture and construal level theory
process and outcome attributes depending on the goal and the infor­
mation content of the attributes. In a process-outcome service attribute According to Hofstede (2001, p. 359), long term orientation “stands
paradigm, customers with low construal level would focus on process for the fostering of virtues oriented towards future rewards, in particular
attributes, which in turn are more relatable to the ‘how’ questions, while perseverance and thrift.” Culture with long term orientation focuses
customers with a high construal level, would focus on outcome attri­ more on future consequences which may lead to behavioral impact on
butes, which are more relatable to the ‘what’ and ‘why’ questions the consumers from such culture. Researchers suggest that ‘by directing
(Tatavarthy, Chatterjee, & Sharma, 2019). We have used the above attention to either the immediate or delayed consequences of one’s ac­
observations on the construal level theory on customer evaluations in tions, consideration of future consequences influences the way in­
order to theorize and explain customer behavior in the context of airline dividuals construe their behavioral options’ (Joireman, Strathman, &
travel. Balliet, 2006, p. 90). Wong and Wyer (2016b) suggested that persons
with a long-term temporal orientation consider future events more
3.2. Travel goals and construal level theory concrete than short-term oriented people. However, their consideration
of present events is more abstract than short-term oriented people. This
Travel goals, namely business vs. leisure, have important role to play means that long-term oriented people are in sync with a higher construal
in consumer evaluations. For instance, the travel goals of business level, which is later supported empirically by van Beek, Handgraaf, and
travelers are more related to the success of the business than the Antonides (2017). As consumers in higher construal level make

3
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

generous evaluation of the services, it is expected that consumers from 3.5. Uncertainty avoidance and expectancy disconfirmation theory
countries with long-term orientation will give higher overall rating and
recommendations for an airline (Huang et al., 2016). Moreover, Extant literature has defined uncertainty avoidance as the extent to
extending the findings of the effect of construal level theory of customer which an individual feels threatened by uncertain or unknown situations
evaluations (Tatavarthy et al., 2019), we also suggest that customers (Hofstede, 1991). An individual with higher degree of uncertainty
with long-term orientation will provide higher importance to abstract avoidance will find it difficult to cope with ambiguity and risks (Hof­
outcome attributes than more vivid process attributes in comparison to a stede, 1991). In a purchase context, in order to avoid ambiguity and
customer with a short-term orientation, as long-term orientation is more risks, consumers with high uncertainty avoidance characteristic actively
associated with a higher construal level. prevent uncertainty by planning and risk aversion (Nakata & Sivakumar,
1996); they take time and collect sufficient information before choosing
H3. Customers belonging from cultures with long-term orientation
purchase. Because of these planning and risk aversion, they are expected
(high construal) are more prone towards (a) positive ratings and (b)
to have a higher level of quality expectations (Donthu & Yoo, 1998).
higher recommendation, than customers belonging from cultures with
Based on the expectation disconfirmation theory, customer satisfaction
short-term orientation (lower construal).
is a function of the difference between the perceived service quality and
H4. Customers belonging from cultures with long-term orientation the expected service quality (Lankton & McKnight, 2012; Tse, Nicosia, &
(high construal) gives (a) higher importance to outcome attributes and Wilton, 1990). Therefore, for a similar level of service perceptions,
(b) lesser importance to process attributes than customers belonging customers with a lower level of service expectations will have higher
from cultures with short-term orientation (lower construal). customer satisfaction and recommendation behavior. Thus, customers
with higher uncertainty avoidance would have higher quality expecta­
3.4. Individualism-collectivism and construal level theory tions, and would thereby provide a lower level of overall rating and
recommendation behavior. Therefore we posit:
While customers who are high in ‘individualism’ recognize them­
H7. Customers belonging from cultures with higher uncertainly
selves as an essential unit and provide priority to themselves, collec­
avoidance are more prone towards (a) lower positive rating and (b)
tivists see themselves as a part of a bigger group and would be
lower recommendation
prioritizing collective goals rather than personal goals (Hofstede, 1980;
Patterson, Cowley, & Prasongsukarn, 2006). The collectivists will have For uncertainty avoiding customers, all the aspects of service quality
higher chances of getting influenced by social perceptions, as they are are essential. However, often they are not able to judge the actual
highly concerned about their social-presentation; they try to avoid quality of service, and create perceptions based on visible or tangible
conflicts, and are less prone to complaining behavior (Patterson et al., cues (Furrer, Liu, & Sudharshan, 2000). Extant literature has provided
2006). Customers who are high in individualism would construe service enough support for this argument; for instance, Donthu and Yoo (1998)
experiences more close to them than customers who are high in argued that uncertainty-avoiding customers use tangibles as a proxy of
collectivism (Wong & Wyer, 2016b). Such lower psychological distance service quality, as these features are visible and helpful to reduce
of the service experience could lead individualistic customers to be in perceived risk associated with a purchase, more so in case of less
lower construal level than collectivist customers (Wong & Wyer, frequent service situations (Furrer et al., 2000). Customers who have a
2016b). Wong and Wyer (2016a, 2016b) also suggested that collectivists higher level of uncertainty avoidance would give more importance to
are expected to construe events related to others more concretely than tangible attributes than intangible attributes of a service per se.
individualists; however, when it comes to activities related to them­
H8. Customers belonging from cultures with a higher level of uncer­
selves, their mental construal levels would be more abstract in com­
tainty avoidance give (a) more importance to tangible attributes and (b)
parison to individualists. Thus collectivists associate self-experiences
lesser importance to intangible attributes.
more with a higher construal level. Importantly, as customers in higher
construal level are more generous during their service evaluations
(Huang et al., 2016), we posit: 3.6. Economy class and expectancy disconfirmation theory

H5. Customers belonging from cultures with low individualism (high

Researchers have observed that middle-income customer groups
construal) are more prone towards (a) positive rating and (b) higher
have the lowest quality expectations from professional services
recommendation than customers belonging from cultures with high
(Webster, 1989). As income grows from mid-income segment to the
individualism (low construal).
high-income segment, the quality expectations grow with it (Webster,
Collectivist contexts have shown the increased importance of the 1989). Based on the expectation disconfirmation theory (Lankton &
needs of a group, or group over self (Lam, Schaubroeck, & Aryee, 2002). McKnight, 2012; Tse et al., 1990), customer satisfaction and overall
The importance of distributive justice therefore, is expected to be higher ratings for economy class travelers, would be much higher than business
in collectivist societies (Pillai, Williams and Justin, 2001); and while the class travelers, leading thereby to higher recommendation behavior.
process attributes of service are more related to procedural justice, the Moreover, for the middle-income segment, the most critical service
outcome attributes are more associated with distributive justice. aspect is value for money. The value of service is also obtained from core
Therefore, collectivist customers give higher importance to outcome service aspects and not the augmented service aspects such as food and
attributes than individualistic customers. Moreover, following the ar­ beverages, in-flight entertainment, etc. Therefore we posit:
guments of individualism-collectivism and construal level, we argue that
H9. Economy class travelers give better (a) overall rating and (b)
collectivists (high construal) would focus more on outcome attributes
recommends more than business class travelers
than individualists (low construal). Therefore we posit:
H10. Economy class travelers give (a) more importance to value for
H6. Customers belonging from cultures with low individualism (high
money and (b) less importance to augmented service aspects.
construal) give (a) higher importance to outcome attributes and (b)
lower importance to process attributes than customers belonging from
4. Methodology
cultures with high individualism (low construal).

4.1. Data and data processing

Online data source is one of the major information sources about

4
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

customer’s service evaluation. Extant literature has shown that big data Table 2
captured from online platforms can be effectively used to study con­ Descriptive statistics of the variables considered.
sumer perception and behavior (Ramanathan, 2010). We have chosen Scale Min Max Mean Standard
online customer reviews given to airlines, as the field data to test the Deviation
hypotheses. We collected data from airlinequality.com, a website, which Recommendation Dummy 0 1 0.63 0.48
gathers airline reviews from reviewers of different countries. Extant behavior Variable
literature has used airlinequality.com as data source for analysis of Overall Rating 10 point 1 10 6.18 3.13
customer word of mouth in online context due to data richness (Chat­ Likert
Seat comfort rating 5 point 1 5 3.29 1.34
terjee, 2019; Siering, Deokar, & Janze, 2018). The website is maintained Cabin Staff Behavior Likert 1 5 3.55 1.44
by Skytrax, an airline service quality certification agency. The website Rating
invites customer reviews in both qualitative and quantitative way on Food and Beverages 1 5 3.06 1.50
various airlines, airports, loungers and seats. The scale, the popularity Service Rating
In-flight Entertainment 1 5 2.57 1.67
and the data variety makes airlinequality.com an apt source of data for
Rating
our study. The data that we used contained customer reviews from the Value for Money rating 1 5 3.46 1.42
period of 2015–2019 where customers have given an overall rating in a Type of traveler Leisure Dummy 0 1 0.16 0.13
1 to 10 point scale (1 means very bad and 10 means very good), and had Type of traveler Variable 0 1 0.73 0.20
also shown their recommendation decisions (RD), i.e., whether to Unknown
Cabin Flown Economy 0 1 0.76 0.18
recommend the airline or not. We have used these variables as our Cabin Flown Unknown 0 1 0.05 0.05
dependent variables. Reviewers have also given ratings on service at­ Individualism Hofstede 1 55 66.84 24.57
tributes such as seat comfort, cabin staff behavior, food and beverages Measures
services, inflight entertainment, ground service, wifi connectivity, and Uncertainty Avoidance Hofstede 1 50 52.99 18.06
Measures
value for money. As ground service (83%) and wifi connectivity (88%)
Long Term Orientation Hofstede 1 62 43.24 21.15
had a lot of missing data, we dropped these variables from our analysis. Measures
The data also contained the origin country of the reviewer, the class of
the travel (economy- 76%, business- 19% or not available- 5%) and the
type of travel (leisure- 16%, business- 11% or not available- 73). We uncertainty avoidance and time-orientation from Hofstede’s measure
have considered ‘not available’ as an important category in the class of (Hofstede, Hofstede, & Minkov, 2005). We have the country of the
travel, and the nature of the trip, as reviewers may not want to disclose reviewer with the country in the dataset mentioned above to find the
these, yet their data could possibly be relevant. We got 28,341 complete individual’s score of these three culture-specific aspects. This method
customer reviews from 132 countries on 345 airlines. We have first also created missing data, as the culture-specific scores were not avail­
chosen 345 airlines randomly from the list of all airlines listed in the able for all countries and all dimensions. This led to a dataset with 20,
website. Then, for each airline, we have randomly chosen a number 467 complete data points. We understand that the above method has
between 100 and 1000 (or maximum number of reviews available for limitations as Hofstede measure should not be used in individual level.
the airline, whichever is lower); that many reviews were collected from The same has been discussed in the limitation section. Moreover, data
the website for the airline. Such randomness in the data collection for all countries was not available, which possibly leads to self-selection
procedure was adopted to make the data as representative as possible. bias. Yet, till now, Hofstede measures are most common measure of
25% of the reviews were from UK, 21.4% were from USA, 12.7% are cultural factors, and the data size is still big enough even with the above
from Australia, 8% from Canada and 3% from Germany. It may be added biases, suggesting us to go ahead with the analysis.
in here that we see higher representation of English speaking countries We changed all independent variables to their standardized form for
because the website is in English. However, as we have not done country our analysis. Table 2 gives the basic descriptive statistics of the variables
level analysis, and used cultural measures for each country instead, low considered in our analysis before they are changed to their standardized
representation of non-English speaking countries did not impact the form.
results. Additionally, there was representation from top 30 airlines of
the world, varying from 1% to 2.4%, suggesting thereby that the data 4.2. Pre-test for process-outcome and tangible-intangible service aspects
had been fairly distributed across airlines.
We found the country-wise scores of individualism-collectivism, We conducted a pre-test to categorize the service attributes (i.e. seat

Table 3
Correlation Table of numeric independent variables.
Seat Cabin Staff Food and In-flight Value for Individualism Uncertainty Long Term
comfort Behavior Rating Beverages Service Entertainment Money Avoidance Orientation
rating Rating Rating rating

Seat comfort rating 1.00

Cabin Staff Behavior 0.41 1.00
Rating ***
Food and Beverages 0.56 0.43*** 1.00
Service Rating ***
In-flight 0.43 0.37*** 0.51*** 1.00
Entertainment ***
Rating
Value for Money 0.51 0.50*** 0.31** 0.40** 1.00
rating ***
Individualism 0.10* 0.06* 0.06* 0.0200 0.08* 1.00
Uncertainty 0.01* 0.01* 0.01* 0.01* 0.00 0.08* 1.00
Avoidance
Long Term 0.05* 0.02* 0.05* 0.01* 0.03* 0.49*** 0.12* 1.00
Orientation

*** means p < 0.001, ** means p < 0.01, * means p < 0.05, “means p < 0.1

5
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

comfort, cabin staff behavior, food and beverages services, inflight effects of the service aspects and travel context (travel type and cabin
entertainment and value for money) into process-outcome and tangible- flown) on recommendation behavior and overall rating respectively.
intangible attributes. For process-outcome, we asked 68 respondents These models have been used to test H1 and H9. Model 1.2 and 2.2 also
(Male ¼ 50%, Female ¼ 50%) to rate whether a certain service attribute include the interaction effects of service aspects and travel contexts,
has focused on the outcome or ‘what’ aspect or is it focusing on the along with the direct effects as mentioned above; these models have
process to reach the outcome or ‘how’ aspect, following extant literature been used to test H2 and H10. Model 1.3 and Model 2.3 include the
(Tatavarthy et al., 2019). We collected the rating in a 1 to 5 point scale direct effects of service aspects and travel context along with the direct
(1 meaning strongly disagree, and 5 meaning strongly agree). We found effects of the cultural parameters of the customers (i.e. individualism,
that seat comfort (Mprocess ¼ 2.15, Moutcome ¼ 4.11, p < 0.01), food and uncertainty avoidance, and long term orientation); these models have
beverages services (Mprocess ¼ 2.77, Moutcome ¼ 3.98, p < 0.05) and value been used to test H3, H5 and H7. Finally, model 1.4 and 2.4 focus on the
for money (Mprocess ¼ 1.87, Moutcome ¼ 3.86, p < 0.01) are outcome direct effects of service aspects, travel context and the cultural param­
aspects, while cabin staff behavior (Mprocess ¼ 4.26, Moutcome ¼ 2.12, p < eters of the customers along with the interaction effects of service as­
0.01) and inflight entertainment (Mprocess ¼ 3.68, Moutcome ¼ 2.25, p < pects and travel contexts and that of services aspects and cultural
0.05) are process aspects. parameters of the customers. These models have been used to test H4,
Similarly, we also asked the respondents whether they can physically H6 and H8. Fig. 1 gives a better explanation of the models, while Table 3
assess the service attributes or whether they think the service attributes documents the results from the regression mentioned above. We found
are more abstract, and cannot be physically accessed. We have taken the the variance inflation factor of the variables in the models 1.1, 1.3, 2.1
measurements on a similar scale as before. We found seat comfort and 2.3 to be less than 2, suggesting no multicollinearity problem in the
(Mtangible ¼ 4.15, Mintangible ¼ 2.23, p < 0.01) and inflight entertainment data.
(Mtangible ¼ 3.73, Mintangible ¼ 2.58, p < 0.05) as tangible, and cabin staff
behavior (Mtangible ¼ 2.11, Mintangible ¼ 4.16, p < 0.01), value for money 5. Results
(Mtangible ¼ 2.32, Mintangible ¼ 3.84, p < 0.05) and food and beverage
services (Mtangible ¼ 2.45, Mintangible ¼ 3.66, p < 0.05) as intangible. We report the result of our analysis in Table 4, where we list down
the coefficients of the regression models. In model 1.1 and 2.1, we
focused on the direct effect of service aspects and consumption contexts
4.3. Analysis
on recommendation behavior and overall rating respectively. We found
all the service aspects to have a positive relationships with recommen­
We used linear regression for explaining the overall rating, and lo­
dation behavior and overall rating, with value for money (recommen­
gistic regression to explain recommendation behavior. Though overall
dation behavior: β ¼ 1.29, p < 0.001; overall rating: β ¼ 0.97, p < 0.001)
rating is an ordinal variable and non-normal in nature, we performed
being the most important followed by cabin staff behavior (recom­
linear regression following extant literature and due to the large sample
mendation behavior: β ¼ 0.83, p < 0.001; overall rating: β ¼ 0.66, p <
size (Siering et al., 2018; Chatterjee, 2019). We also included the results
0.001) and seat comfort (recommendation behavior: β ¼ 0.54, p <
of ordinal logit for overall rating and results are similar. We, at first
0.001; overall rating: β ¼ 0.43, p < 0.001). Food and beverages services
tested the possible correlation amount of the independent variables in
(recommendation behavior: β ¼ 0.21, p < 0.001; overall rating: β ¼
order to explain overall rating and recommendation behavior; Table 3
0.17, p < 0.001) have also been found to be important, while in-flight
documents the same. We did not find a strong correlation among the
entertainment (recommendation behavior: β ¼ 0.064, p < 0.001; over­
independent variables (<0.7), suggesting thereby non-existence of
all rating: β ¼ 0.03, p < 0.001) has been the least important among the
multicollinearity (Chatterjee, 2019).
service aspects.
We used a series of regression models, each building on the findings
We also found that leisure travelers are less prone to
from the previous model. Model 1.1 and 2.1 only include the direct

Fig. 1. Explanation of empirical models.

6
 Table 4
Results from regression analysis.
Model No 1.1 1.2 1.3 1.4 2.1 2.2 2.3 2.4 3.1 3.2 3.3 3.4

Dependent variable Recommendation Recommendation Recommendation Recommendation Overall Overall Overall Overall Overall Overall Overall Overall
behavior behavior behavior behavior Rating Rating Rating Rating Rating Rating Rating Rating
Method Logistic Regression Linear Regression Ordered Logistic Regression
S. Chatterjee and P. Mandal

Sample Size 28,341 28,341 20,467 20,467 28,341 28,341 20,467 20,467 28,341 28,341 20,467 20,467
AIC value 13,002 12,962 9172 9161 83,348 83,042 59,188 58,905
Adj R2 0.7889 0.79 0.7981 0.7997
Coefficients
(Intercept) 8.6*** 9.02*** 8.97*** 9.89*** 1.82*** 1.95*** 1.84*** 2.00***
Seat comfort rating 0.54*** 0.57* 0.54*** 0.77* 0.43*** 0.52*** 0.40*** 0.51** 0.62*** 0.68*** 0.59*** 0.67***
Cabin Staff Behavior 0.83*** 0.62** 0.82*** 0.4900 0.66*** 0.42*** 0.64*** 0.49*** 0.86*** 0.52*** 0.84*** 0.64***
Rating
Food and Beverages 0.21*** 0.69** 0.18*** 1.05** 0.17*** 0.45*** 0.17*** 0.45*** 0.27*** 0.61*** 0.24*** 0.64***
Service Rating
In-flight Entertainment 0.06*** 0.48* 0.08*** 0.18 0.03*** 0.14 0.05*** 0.05 0.07*** 0.22 0.09*** 0.14**
Rating
Value for Money rating 1.29*** 0.77** 1.30*** 0.76* 0.97*** 0.81*** 0.99*** 0.82*** 1.15*** 0.89*** 1.18*** 0.83***
Type of traveler Leisure 0.69* 4.01** 0.50 3.06* 0.08 0.16 0.06 0.20 0.01 0.29 0.11 0.46***
Type of traveler Unknown 0.58* 0.88 0.42 0.75 0.21* 0.21 0.21* 0.01 0.15 0.01 0.09 0.15
Cabin Flown Economy 0.09 1.06** 0.08 1.44*** 0.05* 0.22** 0.02 0.30*** 0.10** 0.03 0.07* 0.09
Cabin Flown Unknown 0.15 0.71 0.16 0.21 0.15** 0.27* 0.17** 0.53*** 0.10 1.41*** 0.08 1.09***
Individualism 0.003 ** 0.01 0.002 0.01 ** 0 0.01 ***
***
Uncertainty Avoidance 0.001* 0.0100 0.002 ** 0.01 0.001 ** 0.01
** ***
Long Term Orientation 0.01*** 0.01 0.002** 0** 0.004* 0.01***

7
Seat comfort rating x 0.1 0.25 0.13 0.17 0.08 0.12*
Type of traveler Leisure
Seat comfort rating x 0.04 0.07 0.09 0.04 0.05 0.09
Type of traveler
Unknown
Seat comfort rating x 0.01 0.08 0.01 0.04 0.01 0.04
Cabin Flown Economy
Seat comfort rating x 0.01 0.12 0.1100 0.16** 0.0800 0.16
Cabin Flown Unknown
Seat comfort rating x 0 0 0
Individualism
Seat comfort rating x 0.001* 0.001* 0.001*
Uncertainty Avoidance
00
Seat comfort rating x 0 0 0
Long Term Orientation
Cabin Staff Behavior 0.4700 0.37 0.08 0.11 0.07 0.01
Rating x Type of
traveler Leisure
Cabin Staff Behavior 0.16 0.14 0.21* 0.23* 0.23* 0.21***
Rating x Type of
traveler Unknown
Cabin Staff Behavior 0.04 0.04 0.0500 0.02 0.11** 0.01*
Rating x Cabin Flown
Economy
Cabin Staff Behavior 0.16 0.06 0.01 0.00 0.33*** 0.3**
Rating x Cabin Flown
Unknown
Cabin Staff Behavior 0 0.003* 0
Rating x Individualism
(continued on next page)
Tourism Management 80 (2020) 104108
 Table 4 (continued )
Model No 1.1 1.2 1.3 1.4 2.1 2.2 2.3 2.4 3.1 3.2 3.3 3.4

Cabin Staff Behavior 0 0 0

Rating x Uncertainty
Avoidance
Cabin Staff Behavior 0.003 * 0 0
S. Chatterjee and P. Mandal

Rating x Long Term

Orientation
Food and Beverages 0.01 0.02 0.07 0.17 0.09 0.18**
Service Rating x Type of
traveler Leisure
Food and Beverages 0.3 0.44 0.1 0.06 0.1 0.04
Service Rating x Type of
traveler Unknown
Food and Beverages 0.20*** 0.21*** 0.22*** 0.22*** 0.32*** 0.33***
Service Rating x Cabin
Flown Economy
Food and Beverages 0.18 0.29 0.22*** 0.29*** 0.42*** 0.44***
Service Rating x Cabin
Flown Unknown
Food and Beverages 0 0 0
Service Rating x
Individualism
Food and Beverages 0 0 0
Service Rating x
Uncertainty Avoidance
Food and Beverages 0.002 * 0 0
Service Rating x Long

8
Term Orientation
In-flight Entertainment 0.11 0.05 0.04 0.05 0.06 0.05
Rating x Type of
traveler Leisure
In-flight Entertainment 0.29 0.19 0.07 0.01 0.08 0.04
Rating x Type of
traveler Unknown
In-flight Entertainment 0.14** 0.1000 0.05** 0.03 0.08*** 0.06*
Rating x Cabin Flown
Economy
In-flight Entertainment 0.42** 0.2700 0.10* 0.06 0.18** 0.09
Rating x Cabin Flown
Unknown
In-flight Entertainment 0.002* 0 0.00100
Rating x Individualism
In-flight Entertainment 0 0 0
Rating x Uncertainty
Avoidance
In-flight Entertainment 0 0 0
Rating x Long Term
Orientation
Value for Money rating x 0.73* 0.74* 0.01 0.07 0.04 0.06
Type of traveler Leisure
Value for Money rating x 0.52* 0.63* 0.02 0.09 0.01 0.05
Type of traveler
Unknown
Value for Money rating x 0.04 0.02 0.16*** 0.19*** 0.29*** 0.32***
Cabin Flown Economy
00
0.4* 0.35 0.33*** 0.31* 0.68*** 0.66***
(continued on next page)
Tourism Management 80 (2020) 104108
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

recommendation behavior than business travelers (recommendation

behavior: β ¼ 4.01, p < 0.01; overall rating: β ¼ 0.08, NS); while

0.002 **

11.12***
12.64***
14.17***
00

9.28***
9.98***
5.16***
6.49***
7.51***
8.22***
0.001
economy class travelers are more satisfied than business class travelers
3.4

(recommendation behavior: β ¼ 0.09, NS; overall rating: β ¼ 0.05, p <

0
0.05). As we include the interaction terms of travel type and travel class
with the service aspects in model 1.2 and 2.2, we find that economy class

10.23***
11.36***
12.88***
14.40***
5.47***
6.75***
7.77***
8.47***
9.52***
travelers are more prone to recommendation behavior too (recommen­
3.3

dation behavior: β ¼ 1.06, p < 0.05). The above results support H1 and
H9, which expects business travelers to be more accommodating than
leisure travelers, while business class travelers to be more demanding

11.08***
12.57***
14.05***
9.94***
5.08***
6.43***
7.45***
8.16***
9.22***
than economy class travelers.
3.2

In model 1.3 and 2.3 we have included the cultural components of

the customers in the model and found that customers from countries
with a higher level of individualism will have more negative customer
10.12***
11.26***
12.74***
14.22***
5.31***
6.64***
7.64***
8.35***
9.40***

outcome (recommendation behavior: β ¼ 0.003, p < 0.01; overall

3.1

rating: β ¼ 0.002, p < 0.001); this supports H5. We also found that
customers who are high in uncertainty avoidance would have a higher
satisfaction and recommendation behavior (recommendation behavior:
00
0.002

0.001

β ¼ 0.001, p < 0.05; overall rating: β ¼ 0.002, p < 0.01), while

***
2.4

customers who are long-term oriented, would have higher customer

outcomes (recommendation behavior: β ¼ 0.01, p < 0.001; overall rat­
ing: β ¼ 0.002, p < 0.001); these results support H3 and H7.
In models 1.2 and 2.2, we introduced the interaction effects of travel
2.3

type and travel class with the service aspects covered on model 1.1 and
1.2. We found that cabin staff behavior (process) for instance was more
important to leisure travelers than business travelers (recommendation
behavior: β ¼ 0.47, p < 0.1; as per model 1.4: β ¼ 0.71, p < 0.05), while
2.2

the value for money (outcome) was less important (recommendation

behavior: β ¼ 0.73, p < 0.05). According to H2, business travelers
(high construal) give (a) higher importance to outcome attributes and
(b) and lower importance to process attributes, in comparison with
2.1

leisure travelers (low construal); this therefore supports H2. On the

other hand, for economy class travelers, food and beverage (recom­
mendation behavior: β ¼ 0.20, p < 0.001; overall rating: β ¼ 0.22, p
< 0.001) and in-flight entertainment (recommendation behavior: β ¼
0.14, p < 0.001; overall rating: β ¼ 0.05, p < 0.001) are less critical,
while value for money is more important (overall rating: β ¼ 0.16, p <
0.00100

0.001) than business class travelers. This supports H10, which posits
1.4

that economy class travelers give more importance to value for money
and less importance to augmented service aspects.
In model 1.4 and 2.4, we have included the interaction term of cul­
tural components with service aspects in model 1.3 and 2.3 respectively.
We found that for customers from counties with a higher level of indi­
*** means p < 0.001, ** means p < 0.01, * means p < 0.05, “means p < 0.1

vidualism, cabin staff behavior (overall rating: β ¼ 0.003, p < 0.05) and
1.3

inflight entertainment (recommendation behavior: β ¼ 0.002, p < 0.05)

are more important than value for money (overall rating: β ¼ 0.002, p
< 0.001). This is because in low construal process, attributes are of
higher importance, and outcome attributes are of lower importance; our
results support H6. Customers in long-term orientation are at a higher
construal level, giving lesser importance to process traits such as cabin
staff behavior (recommendation behavior: β ¼ 0.003, p < 0.05), while
1.2

stressing more on outcome attributes such as food and beverages

(recommendation behavior: β ¼ 0.002, p < 0.05), thus supporting H4.
Lastly, we found that for customers with a higher level of uncertainty
avoidance seat comfort, a tangible attribute, is more important
(recommendation behavior: β ¼ 0.001, p < 0.05; overall rating: β ¼
0.001, p < 0.05), supporting in turn H8.
1.1

6. Country wise analysis

Value for Money rating x

Value for Money rating x

Value for Money rating x

Value for Money rating x

Long Term Orientation
Cabin Flown Unknown

Uncertainty Avoidance
Table 4 (continued )

For country wise analysis we have chosen USA, Canada, Germany,

Singapore and India based on data availability and their extremeness in
Individualism

terms of cultural factors. While USA and Canada is high in individu­

Model No

alism, Singapore and Germany is high in long term orientation. More­

9 | 10

over, Singapore is very low in uncertainty avoidance. India, on the other

5|6
6|7
7|8
8|9
1|2
2|3
3|4
4|5

hand, is moderate in all the aspects. Table 5 reports the results for the

9
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

Table 5
Country wise analysis of the preferences.
Country USA Canada Singapore Germany India

IDV 91 80 20 67 48
UAI 46 48 8 65 40
LTR 26 36 72 83 51
Dependent variable RBV OVRT RBV OVRT RBV OVRT RBV OVRT RBV OVRT
Regression Method Logistic Linear Logistic Linear Logistic Linear Logistic Linear Logistic Linear
Sample Size 5732 5732 2658 2658 537 537 887 887 433 433
AIC value 2181 965 301 412 248
Adj R2 0.83 0.82 0.71 0.77 0.74
Coefficients
(Intercept) 9.94 1.81 10.27 1.59 7.65 0.89 8.85 1.58 7.92 0.98
*** *** *** *** *** *** *** *** *** ***
SEAT 0.54 0.37 0.51 0.38 0.31 0.36 0.52 0.43 0.54 0.39
*** *** *** *** “ *** *** *** ** ***
CBNSTF 0.82 0.60 0.80 0.72 0.58 0.45 0.57 0.50 0.63 0.56
*** *** *** *** *** *** *** *** *** ***
FNB 0.14 0.17 0.12 0.15 0.08 0.07 0.01 0.14 0.32 0.15
*** *** *** *** * *
ENT 0.11 0.06 0.23 0.12 0.09 0.04 0.25 0.06 0.10 0.06
*** *** *** *** ** “
VMON 0.91 1.11 1.21 1.18 1.44 1.46 1.12 0.86 1.57 0.94
*** *** *** *** *** *** *** *** *** ***

RBV ¼ Recommendation behavior, OVRT¼ Overall Rating, SEAT¼ Seat comfort rating, CBNSTF ¼ Cabin Staff Behavior Rating, FNB¼ Food and Beverages Service
Rating, ENT ¼ In-flight Entertainment Rating, VMON ¼ Value for Money rating, TTL ¼ Type of traveler Leisure, TTNA ¼ Type of traveler Unknown, CFE¼ Cabin Flown
Economy, CFNA ¼ Cabin Flown Unknown, IDV ¼ Individualism, UAI ¼ Uncertainty Avoidance, LTR ¼ Long Term Orientation

country wise analysis. associated with lower construal levels, as their goals are ‘then and there’
oriented, and thereby more related to service aspects (Huang et al.,
6.1. Individualism 2016; Kashyap & Bojanic, 2000). Following the construal level theory,
they would be less satisfied and would thereby recommend less (Eyal
For consumers from countries with higher individualism (USA, et al., 2004; Huang et al., 2016). Moreover, they would tend to give
Canada), the intercepts are lower than consumers from countries with more importance to process attributes. As customers belonging from
lower individualism (Singapore, Germany, India), suggesting that cus­ cultures, which are long-term oriented, tend to be closely associated
tomers belonging from cultures with low individualism (high construal) with a higher construal level (Joireman et al., 2006; van Beek et al.,
are more prone towards (a) positive rating and (b) higher recommen­ 2017; Wong & Wyer, 2016b), it is expected that they would give higher
dation than customers belonging from cultures with high individualism importance to outcome attributes (Tatavarthy et al., 2019), and would
(low construal), thus supporting H5. also be more accommodating during service evaluation and spreading
Moreover, consumers from countries with higher individualism word of mouth (WoM) (Eyal et al., 2004; Huang et al., 2016), as seen in
(USA, Canada) has higher co-efficient values for cabin-staff behavior our study. Customers with a higher level of individualism are closely
and in-flight entertainment (process attributes) and lower coefficient associated with lower construal levels (Wong and Wyer, 2016a, 2016b),
value for value for money (outcome attribute) than consumers from leading to a stricter evaluation of services (Eyal et al., 2004; Huang et al.,
countries with lower individualism (Singapore, Germany, India). This 2016), and higher importance to process attributes. While process at­
supports H6. tributes are more associated with procedural justice and outcomes with
distributive justice, it may be noted that collectivist societies focus more
on distributive justice, while individualist societies focus more on pro­
6.2. Long term orientation
cedural justice (Lam et al., 2002; Pillai, Williams, & Justin Tan, 2001).
Thus, the higher importance of process is expected as also seen in our
Consumers from countries with higher long term orientation
results.
(Singapore, Germany) has lower importance of food and beverages, an
Second, customers with a high level of uncertainty avoidance will try
outcome attribute, towards their consumer outcomes. This supports H4.
to reduce their risks by gathering enough information before purchase
and not compromising with quality (Nakata & Sivakumar, 1996).
6.3. Uncertainty avoidance Therefore, their quality expectations are also higher (Donthu & Yoo,
1998). Moreover, middle-income customers, who generally travel in
Consumers from counties with lower uncertainty avoidance economy class, have the lowest quality expectations (Webster, 1989).
(Singapore) has lower importance of tangible attributes such as seat Based on the expectation disconfirmation theory, customer satisfaction
comfort and inflight entertainment, in comparison to other counties. rises as quality expectations drop (Lankton & McKnight, 2012; Tse et al.,
This supports H8. 1990), which in turn leads to higher recommendation behavior; the
situation reverses, when quality expectations rise. We also show that
7. Discussions middle-income segment gets higher utility from value for money and
lower utility from augmented service aspects such as food and bever­
Our study focused on how travel-goals along with economic and ages, in-flight entertainment, etc.
cultural background of customers impact their satisfaction levels and Lastly, though uncertainty avoiding customers are risk averse, they
recommendation behavior. We have used the construal level theory are often not experts in service evaluations, and make perceptions based
coupled with the expectation disconfirmation theory in order to explain on visible or tangible cues, more so in less frequent service situations
the results. such as air travel (Donthu & Yoo, 1998; Furrer et al., 2000). Therefore,
Here, we focus on how the construal level theory plays a role in the customers with a higher level of uncertainty avoidance will give more
service evaluation process. We suggest that leisure travelers are closely

10
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

importance to tangible attributes than intangible attributes of a service, tangible-intangible and process-outcome dimensions. Within a sophis­
as shown in our results. ticated service setting like the ‘airlines’, where the overall service
quality is dependent on multiple parallels and sequential service con­
texts, understanding about such drivers is crucial.
7.1. Theoretical contribution
Second, the current study helps airline managers to design a service
offering based on the customer’s travel goal, socio-economic and cul­
The theoretical contribution of this study is manifold. First, this study
tural background. Such necessary information is often collected during
focused on the role of culture in forming customer preferences and a
the booking of an airline seat, and the managers can use such informa­
service evaluation process, in the context of the airlines industries.
tion to generate customized packages, communication strategies, cam­
While existing literature has evaluated the role of culture in service
paigns, etc. to make the customers more satisfied, thereby encouraging
evaluation, especially in the context of ‘Western’ and Asian cultures
positive word of mouth.
(Mattila, 1999), studies in the context of airlines are scarce. The
Third, methodologically speaking, the study highlights how user-
uniqueness of our study is that we attempt to provide a cross-cultural
generated content could be used for customer segmentation, more so
exploration of the drivers of the overall rating and recommendation
in the post-purchase context; and how dataset from multiple sources
behavior in the context of airlines, using user-generated data on
could be merged to create insights about customers. Here, we combined
customer reviews. Thus, we contribute to extant literature on
user-generated content with Hofstede’s cultural dimension data to
cross-cultural studies on customer behavior, specifically in the airlines
develop actionable insights. Such a methodology is novel in the context
context (Crotts & Erdmann, 2000; Ruiz-Mafe et al., 2013).
of airline customer relationship management.
Second, the interaction of construal level and cultural dimensions
have not been well explored yet. Wong and Wyer (2016a, 2016b) have
found that individualistic or short-term oriented individuals behave in 7.3. Limitations and future scope
sync with low construal level (Trope & Liberman, 2010), while collec­
tivists or long-term oriented individuals behave differently; this The study is using country-level cultural measures for individual
behavioral shift could possibly explain the differential results. Kim, level analysis. People within a same culture can have different types of
Sung, and Drumwright (2018) found that the construal level of cus­ personality traits, which possibly cannot be measured, as the same may
tomers are often driven by their culture of origin. In this study, while not available as a secondary data source. Future researchers could
explaining cross-cultural differences along with the importance of thereby choose more suitable or alternative measures to validate our
various drivers towards the overall rating and loyalty systems, we used results. Moreover, the pre-test sample could also be larger and more
the construal level theory, and found that cultural dimensions, such as diverse in order to cater to possible effects of culture on tangible-
long-term orientation, individualism-collectivism, etc. are indeed asso­ intangible and process-outcome categorization of service attributes;
ciated with the construal level. Such a finding is a noble contribution in nevertheless, we did try to overcome this limitation by reporting cross-
extant literature on both customer culture (Kim et al., 2018; Wong and country comparisons. Interestingly, the results of cross-country com­
Wyer, 2016a, 2016b) and construal level theory, especially in the parisons also support our hypotheses. The other salient limitation would
context of service evaluation (Trope and Liberman, 2010; Ding & Keh, be that we focused only on quantitative ratings and not on textual re­
2017; Hern� andez-Ortega, 2018). views, which is also a vital source of information. Future research could
Third, the study also contributes to extant literature on customer focus on data retrieved from textual reviews given by customers in post-
segmentation in the context of the airline Industry. Extant research has purchase situations in order to create insights about customer prefer­
used survey-based data followed by machine learning or Bayesian sta­ ences and evaluations. Another limitation could be the inclusion of other
tistical techniques to form segments in the airline industry (Marti­ cultural dimensions, which had to be dropped due to multicollinearity
nez-Garcia & Royo-Vela, 2010; Teichert et al., 2008; Wen & Lai, 2010). issues. One way of handling such a problem could be by performing a
While the survey method has its limitations, none of the studies could principal component analysis of the cultural dimensions before actually
give the underlying psychological process, which actually leads to dif­ using them in the regression models. Future research could look to find
ferential preferences of customers under different segments. Moreover, novel ways of focusing on all cultural aspects in the post-purchase
all of them focused on the purchase choice and not on post-purchase customer evaluation context using real customer data. We strongly
evaluation, which is also crucial, especially in the context of industries suggest qualitative and mixed-method based studies including content
like the ‘airlines’ that relies on loyalty and positive word of mouth analysis of the customer reviews to create further detailed insight from
(WoM). We focused on the customers’ overall rating and recommen­ the data.
dation behavior based on their post-purchase reviews, whereby we
found answers to the ‘how’ and ‘why’, which in essence are the drivers of Declaration of competing interest
overall rating and recommendation behavior, and differ from one
traveler to the other based on their class of travel and goal of travel. None.
Thus, this study contributes to extant literature on customer segmen­
tation in the context of the airlines industry (Martinez-Garcia & CRediT authorship contribution statement
Royo-Vela, 2010; Teichert et al., 2008; Wen & Lai, 2010).
Swagato Chatterjee: Conceptualization, Methodology, Software,
7.2. Managerial implications Formal analysis, Writing - original draft, Supervision, Funding acquisi­
tion. Prasenjit Mandal: Validation, Resources, Writing - review &
The study has many managerial implications too. First, the study editing, Visualization.
explores the drivers of overall rating and recommendation behavior in
the context of the airlines industry. Such information will help airline Acknowledgement
managers better manage their service attributes to ensure best customer
outcomes. They can prioritize and optimally allocate their resources The authors like to thank Sponsored Research and Industrial Con­
based on the driving strength of various service attributes, categorized in sultancy (SRIC, IIT Kharagpurr for funding the project through Institute

11
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

Scheme for Industrial Research and Development (ISIRD). We also like Kruglanski, A. W. (1975). The endogenous-exogenous partition in attribution theory.
Psychological Review, 82(6), 387–406.
to thank the editorial office of Tourism Management for handling the
Lam, S. S., Schaubroeck, J., & Aryee, S. (2002). Relationship between organizational
review process of the paper. We also like to thank the reviewers whose justice and employee work outcomes: A cross-national study. Journal of
valuable comments have enriched the paper. Organizational Behavior, 23(1), 1–18.
Lankton, N. K., & McKnight, H. D. (2012). Examining two expectation disconfirmation
theory models: Assimilation and asymmetry effects. Journal of the Association for
References Information Systems, 13(2), 1.
Liberman, N., Sagristano, M. D., & Trope, Y. (2002). The effect of temporal distance on
Akamavi, R. K., Mohamed, E., Pellmann, K., & Xu, Y. (2015). Key determinants of level of mental construal. Journal of Experimental Social Psychology, 38(6), 523–534.
passenger loyalty in the low-cost airline business. Tourism Management, 46, 528–545. Liberman, N., Trope, Y., & Wakslak, C. (2007). Construal level theory and customer
Arnould, E. J., & Thompson, C. J. (2005). Consumer culture theory (CCT): Twenty years behavior. Journal of Customer Psychology, 17(2), 113–117.
of research. Journal of Consumer Research, 31(4), 868–882. Lin, Z., & Vlachos, I. (2018). An advanced analytical framework for improving customer
van Beek, J., Handgraaf, M. J., & Antonides, G. (2017). Time orientation and construal satisfaction: A case of air passengers. Transportation Research Part E: Logistics and
level: Effects on eating and exercising behaviour and preferences. International Transportation Review, 114, 185–195.
journal of customer studies, 41(1), 54–60. Lu, J. L. (2017). Segmentation of passengers using full-service and low-cost
Chang, L. Y., & Hung, S. C. (2013). Adoption and loyalty toward low cost carriers: The carriers–Evidence from Taiwan. Journal of Air Transport Management, 62, 204–216.
case of Taipei–Singapore passengers. Transportation Research Part E: Logistics and Martinez-Garcia, E., & Royo-Vela, M. (2010). Segmentation of low-cost flights users at
Transportation Review, 50, 29–36. secondary airports. Journal of Air Transport Management, 16(4), 234–237.
Chatterjee, S. (2019). Explaining customer ratings and recommendations by combining Mason, K. (2002). Future trends in business travel decision making. Journal of Air
qualitative and quantitative user generated contents. Decision Support Systems, 119, Transportation, 7, 47–69, 2002.
14–22. Mason, K. J. (2005). Observations of fundamental changes in the demand for aviation
Cho, W., & Min, D. J. (2018). Longitudinal examination of passenger characteristics services. Journal of Air Transport Management, 11(1), 19–25.
among airline types in the US. Journal of Air Transport Management, 72, 11–19. Mattila, A. S. (1999). The role of culture in the service evaluation process. Journal of
Crotts, J. C., & Erdmann, R. (2000). Does national culture influence customers’ Service Research, 1(3), 250–261.
evaluation of travel services? A test of Hofstede’s model of cross-cultural differences. Nakata, C., & Sivakumar, K. (1996). National culture and new product development: An
Managing Service Quality: International Journal, 10(6), 410–419. integrative review. Journal of Marketing, 60(1), 61–72.
Ding, Y., & Keh, H. T. (2017). Customer reliance on intangible versus tangible attributes Ostrowski, P. L., O’Brien, T. V., & Gordon, G. L. (1993). Service quality and customer
in service evaluation: The role of construal level. Journal of the Academy of Marketing loyalty in the commercial airline industry. Journal of Travel Research, 32(2), 16–24.
Science, 45(6), 1–18. Park, J. H., Gu, B., Leung, A. C. M., & Konana, P. (2014). An investigation of information
Doganis, R. (2006). The airline business. UK: Routledge. sharing and seeking behaviors in online investment communities. Computers in
Donthu, N., & Yoo, B. (1998). Cultural influences on service quality expectations. Journal Human Behavior, 31, 1–12.
of Service Research, 1(2), 178–186. Park, J. W., Robertson, R., & Wu, C. L. (2004). The effect of airline service quality on
Eyal, T., Liberman, N., Trope, Y., & Walther, E. (2004). The pros and cons of temporally passengers’ behavioural intentions: A Korean case study. Journal of Air Transport
near and distant action. Journal of Personality and Social Psychology, 86(6), 781. Management, 10(6), 435–439.
Furrer, O., Liu, B. S. C., & Sudharshan, D. (2000). The relationships between culture and Patterson, P. G., Cowley, E., & Prasongsukarn, K. (2006). Service failure recovery: The
service quality perceptions: Basis for cross-cultural market segmentation and moderating impact of individual-level cultural value orientation on perceptions of
resource allocation. Journal of Service Research, 2(4), 355–371. justice. International Journal of Research in Marketing, 23(3), 263–277.
Gideon, L. (2012). Handbook of survey methodology for the social sciences. New York: Pillai, R., Williams, E. S., & Justin Tan, J. (2001). Are the scales tipped in favor of
Springer. procedural or distributive justice? An investigation of the US, India, Germany, and
Hern�andez-Ortega, B. (2018). Don’t believe strangers: Online customer reviews and the Hong Kong (China). International Journal of Conflict Management, 12(4), 312–332.
role of social psychological distance. Information & Management, 55(1), 31–50. Ramanathan, R. (2010). The moderating roles of risk and efficiency on the relationship
Herzog, S. M., Hansen, J., & W€ anke, M. (2007). Temporal distance and ease of retrieval. between logistics performance and customer loyalty in e-commerce. Transportation
Journal of Experimental Social Psychology, 43(3), 483–488. Research Part E: Logistics and Transportation Review, 46(6), 950–962.
Higgins, E. T., & Trope, Y. (1990). Activity engagement theory: Implications of multiply Reichheld, F. F. (2003). The one number you need to grow. Harvard Business Review, 81
identifiable input for intrinsic motivation. In E. T. Higgins, & R. M. Sorrentino (Eds.), (12), 46–55.
Handbook of motivation and cognition: Foundations of social behavior (Vol. 2, pp. Reisinger, Y., & Crotts, J. C. (2010). Applying Hofstede’s national culture measures in
229–264). New York: Guilford Press. tourism research: Illuminating issues of divergence and convergence. Journal of
Ho-Dac, N. N., Carson, S. J., & Moore, W. L. (2013). The effects of positive and negative Travel Research, 49(2), 153–164.
online customer reviews: Do brand strength and category maturity matter? Journal of Ruiz-Mafe, C., Sanz-Blas, S., Hernandez-Ortega, B., & Brethouwer, M. (2013). Key drivers
Marketing, 77(6), 37–53. of customer purchase of airline tickets: A cross-cultural analysis. Journal of Air
Hofstede, G. (1980). Culture and organizations. International Studies of Management & Transport Management, 27, 11–14.
Organization, 10(4), 15–41. Siering, M., Deokar, A. V., & Janze, C. (2018). Disentangling consumer
Hofstede, G. (1991). Organizations and cultures: Software of the mind. New York: recommendations: Explaining and predicting airline recommendations based on
McGrawHill. online reviews. Decision Support Systems, 107, 52–63.
Hofstede, G. (2001). Culture’s consequences: Comparing values, behaviors, institutions and Steenkamp, J. B. E. (2019). Global versus local consumer culture: Theory, measurement,
organizations across nations. Sage publications. and future research directions. Journal of International Marketing, 27(1), 1–19.
Hofstede, G., Hofstede, G. J., & Minkov, M. (2005). Cultures and organizations: Software of Tatavarthy, A. D., Chatterjee, S., & Sharma, P. (2019). Exploring the moderating role of
the mind (Vol. 2). New York: Mcgraw-hill. construal levels on the impact of process versus outcome attributes on service
Huang, N., Burtch, G., Hong, Y., & Polman, E. (2016). Effects of multiple psychological evaluations. Journal of Service Theory and Practice, 30, 1–40.
distances on construal and customer evaluation: A field study of online reviews. Teichert, T., Shehu, E., & von Wartburg, I. (2008). Customer segmentation revisited: The
Journal of Customer Psychology, 26(4), 474–482. case of the airline industry. Transportation Research Part A: Policy and Practice, 42(1),
Huse, C., & Evangelho, F. (2007). Investigating business traveller heterogeneity: Low- 227–242.
cost vs full-service airline users? Transportation Research Part E: Logistics and Trope, Y., & Liberman, N. (2010). Construal-level theory of psychological distance.
Transportation Review, 43(3), 259–268. Psychological Review, 117(2), 440.
Hwang, J., & Hyun, S. S. (2017). First-class airline travelers’ tendency to seek Tse, D. K., Nicosia, F. M., & Wilton, P. C. (1990). Customer satisfaction as a process.
uniqueness: How does it influence their purchase of expensive tickets? Journal of Psychology and Marketing, 7(3), 177–193.
Travel & Tourism Marketing, 34(7), 935–947. Vallacher, R. R., & Wegner, D. M. (1987). What do people think they’re doing? Action
Hwang, J., & Lyu, S. O. (2018). Understanding first-class passengers’ luxury value identification and human behavior. Psychological Review, 94(1), 3–15.
perceptions in the US airline industry. Tourism management perspectives, 28, 29–40. Vlachos, I., & Lin, Z. (2014). Drivers of airline loyalty: Evidence from the business
Joireman, J., Strathman, A., & Balliet, D. (2006). Considering future consequences: An travelers in China. Transportation Research Part E: Logistics and Transportation Review,
integrative model. In L. Sanna, & E. Chang (Eds.), Judgments over time: The interplay 71, 1–17.
of thoughts, feelings, and behaviors (pp. 82–99). Oxford, UK: Oxford University Press. Webster, C. (1989). Can customers be segmented on the basis of their service quality
Kashyap, R., & Bojanic, D. C. (2000). A structural analysis of value, quality, and price expectations? Journal of Services Marketing, 3(2), 35–53.
perceptions of business and leisure travelers. Journal of Travel Research, 39(1), Wen, C. H., & Lai, S. C. (2010). Latent class models of international air carrier choice.
45–51. Transportation Research Part E: Logistics and Transportation Review, 46(2), 211–221.
Kim, D. H., Sung, Y., & Drumwright, M. (2018). ‘Where I come from’determines,‘how I Williams, L. E., & Bargh, J. A. (2008). Keeping one’s distance: The influence of spatial
construe my future’: The fit effect of culture, temporal distance, and construal level. distance cues on affect and evaluation. Psychological Science, 19(3), 302–308.
International Journal of Advertising, 37(2), 270–288. Wong, V. C., & Wyer, R. S., Jr. (2016a). Mental traveling along psychological distances:
Kim, S. S., & Prideaux, B. (2003). A cross-cultural study of airline passengers. Annals of The effects of cultural syndromes, perspective flexibility, and construal level. Journal
Tourism Research, 30(2), 489–492. https://doi.org/10.1016/S0160-7383(02)00111- of Personality and Social Psychology, 111(1), 17.
1. Wong, V. C., & Wyer, R. S., Jr. (2016b). Mental traveling along social and temporal
Kim, Y. K., & Lee, H. R. (2009). Passenger complaints under irregular airline distances: The influence of cultural syndromes on construal level. In P. Moreau, &
conditions–cross-cultural study. Journal of Air Transport Management, 15(6), S. Puntoni (Eds.), NA - advances in customer research (Vol. 44, pp. 685–686). Duluth,
350–353. MN: Association for Customer Research.

12
S. Chatterjee and P. Mandal Tourism Management 80 (2020) 104108

Dr. Prasenjit Mandal is an academician and consultant in the

area of information science and operations. He has published in
European Journal of Operations Research and presented in
Dr. Swagato Chatterjee is a researcher, consultant, teacher and various national and international conferences
academician. He has research interests in tourism, marketing,
operations and analytics. He has published in reputed inter­
national journals such as Decision Support Systems, International
Journal of Hospitality Management, Journal of Service Theory and
Practice, Journal of Strategic Marketing, Journal of Indian Busi­
ness Research, Global Business Review among others and pre­
sented in various national and international conferences.

13