TE-published

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/280206550
Culture, Product Differentiation and Market
Segmentation: A Structural Analysis of the
Motivation and Satisfaction of Touri....
Article in Tourism Economics · June 2015
DOI: 10.5367/te.2015.0483
CITATIONS
READS
35
1,269
4 authors, including:
João Romão
Peter Nijkamp
Yasuda Women's University
Tinbergen Institute Amsterdam
58 PUBLICATIONS 588 CITATIONS
1,083 PUBLICATIONS 24,603 CITATIONS
SEE PROFILE
SEE PROFILE
Eveline S. Van Leeuwen
Wageningen University & Research
103 PUBLICATIONS 950 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
AG2020 View project
AG2020 View project
All content following this page was uploaded by João Romão on 23 November 2017.
The user has requested enhancement of the downloaded file.
Tourism Economics, 2015, 21 (3), 455–474 doi: 10.5367/te.2015.0483
Culture, product differentiation and market
segmentation: a structural analysis of the
motivation and satisfaction of tourists
in Amsterdam
JOÃO ROMÃO
Research Centre for Spatial and Organizational Dynamics, University of Algarve, Faro,
Portugal. E-mail: [email protected]. (Corresponding author.)
BART NEUTS
School of Hospitality and Tourism, AUT University, Auckland, New Zealand.
PETER NIJKAMP
Department of Spatial Economics, Faculty of Economics and Business Administration,
VU University, Amsterdam, the Netherlands.
EVELINE VAN LEEUWEN
Department of Spatial Economics, Faculty of Economics and Business Administration,
VU University, Amsterdam, the Netherlands.
The varied supply of tourism services – with particular emphasis on
tangible and intangible cultural aspects – corresponds ideally to
visitors’ characteristics and wishes. This paper considers a major
tourist destination, such as Amsterdam, as an export-oriented multiproduct company, characterized by spatial and functional market
segmentation and monopolistic competition reflected in product
differentiation. Urban branding and attractiveness may favour tourist
destination loyalty. The complex decision web of motivation and
satisfaction of tourists in Amsterdam is analysed with a structural
equations model (SEM). The authors find that different tourist profiles,
in terms of personal characteristics and motivations, can significantly
impact the satisfaction received from tourism services. Furthermore,
and most interestingly, the results suggest that satisfaction does not
necessarily lead to improved destination loyalty, but is contingent on
the source of satisfaction. In this case, satisfaction resulting from
tangible or intangible cultural sources has clearly different implications
for loyalty, with relevant managerial implications.
Keywords: market segmentation; tourist satisfaction; destination loyalty;
intangible heritage; structural equations model
456
TOURISM ECONOMICS
Even though travelling is an ancient phenomenon, not until the 1960s did
tourism become a widespread leisure activity, increasingly motivated by pleasure
and relaxation. Since that time we have seen a rapid rise in this new leisure
society, with 2012 for the first time registering over one billion international
tourist arrivals (United Nations World Tourism Organization (UNWTO), 2013).
Not only has travelling become more accessible to a broader public; it has also
become increasingly diversified. This development prompts a new challenge to
tourism destinations: can they offer a broad package of tourism facilities to meet
the particular demands of specific groups of tourists? A wide spectrum of
tourism services is able to attract a diversified group of potential clients and
to potentially increase their loyalty, making the tourism region less vulnerable
to various demand side shifts or to the seasonality of tourism.
A tourism destination is not a set of distinct natural, cultural, artistic or
environmental resources, but an inclusive appealing product complex that is
offered in a certain appropriate place; it is based on a broadly composed and
integrated portfolio of services offered by a place or destination that supplies
a multidimensional holiday experience, which meets the various needs of a
heterogeneous group of modern tourists. A tourism destination thus produces
a composite package of tourist services based on its indigenous supply (or
attraction) potential. It should be added that the attractiveness of a city as an
urban tourist complex depends not only on the presence of facilities of all kinds,
but also on the information provided on these facilities. Thus, tourism
marketing has become a critical success factor for each destination. And therefore,
web-based information (such as pre-trip information) and electronic information
devices (such as portable GPS equipment) are also of great importance (Neuts
et al, 2013a), with implications for the satisfaction achieved by tourists at the
destination (Neuts et al, 2013b).
From this perspective, tourism destinations produce a large set of products
and services under the same brand, and the tourist’s overall experience is the
result of the satisfaction received from multiple experiences related to these
products and services (Buhalis, 2000). Consequently, tourism destinations
become heterogeneous multi-product, multi-client business organizations. One
may therefore regard such modern tourist complexes as export-oriented multiproduct companies, characterized by spatial and functional market segmentation,
and by monopolistic competition, reflected in product differentiation (Matias
et al, 2007).
In a competitive environment, product differentiation is therefore of primary
importance to the attraction of new tourists and provision of incentives for
repeat visits. This latter aspect, loyalty to a destination, is commonly assumed
to be an important aspect of destination marketing: it is less costly to attract
a satisfied visitor than a new one; tourists are better informed in the repeat visits
(implying that they can reach higher levels of satisfaction); and they promote
the destination at no cost in a very effective way (word of mouth among their
circle of friends). In fact, repeat visitors can contribute to the achievement of
higher revenues and profits for tourism companies. Loyalty improvement could
then be considered an effective way to increase destination competitiveness and
it therefore is of primary importance to identify the product characteristics that
increase tourist satisfaction and the way in which this satisfaction might
translate into loyalty.
Culture, product differentiation and market segmentation
457
This study therefore aims to conceptualize and model the force field of
modern tourism, from both the demand and supply sides. Particular attention
will be paid to the motivation, satisfaction and loyalty of tourists concerning
the destination of Amsterdam. Amsterdam received 5 million tourists in 2007
(when the data for this article were collected), a number that increased to 5.7
million in 2012, making Amsterdam one of the top 10 European cities for
tourism. The importance of cultural assets for the attractiveness of the city is
clearly expressed by the percentage of tourists who visit at least one museum:
73% in 2007 and 85% in 2011 (ATCB, 2012).
This article pays special attention to the role of cultural heritage in attracting
visitors and to its implications for the satisfaction and loyalty of the tourists,
by presenting an innovative approach, based on a broad concept of urban
cultural heritage including the built environment (museums, monuments or
architecture), cultural events (concerts, festivals and other happenings) and also
intangible cultural aspects (local knowledge, local values or lifestyle), with
tourists being given the opportunity to express their opinion on what was to
be considered cultural heritage during the empirical fieldwork. The paper is
organized as follows. The second section, provides the necessary background
based on a literature search for conceptualizing the structure of a tourism
complex. Then, the third section discusses the database on the city of Amsterdam,
while the fourth section presents the statistical methodology. The fifth and sixth
sections present and interpret the results. In the seventh and final section we
offer our concluding remarks.
Motivation, satisfaction and loyalty in tourism
As mentioned above, the tourism market is a varied and multi-client system.
The heterogeneity of tourists is related to their personal characteristics; their
origin, age, level of education, social conditions, cultural values or other
individual attributes all influence the choice of tourism destinations and the
related expectations, perceptions and motivations they bring to the destination.
The identification of this market heterogeneity is assumed in the literature to
be an extremely important element in defining effective marketing strategies
(Kozak and Rimmington, 2000; Castro et al, 2007). In fact, the place image
created by each tourist about a destination influences his or her decision to
travel, the choice of a destination, the motivations to experience particular
aspects of each place, the choice of products and services to be consumed during
the holiday, satisfaction with the travel, and, consequently, the loyalty to a
destination (Chen and Tsai, 2007).
There is a general consensus in the literature about the importance of
identifying push and pull motivations to define proper marketing strategies.
As proposed by Crompton (1979), push factors influence the decision to travel
and are related to intangible and intrinsic personal preferences of tourists
(relaxation, entertainment, escape from routine, adventure, sports); pull factors
(culture, heritage, museums, climate, landscape) affect the choice of a specific
destination, and are related to the tangible attributes of each place (Dann, 1981;
Kozak, 2002; Bansall and Eizelt, 2004; Yoon and Uysal, 2005). As the push
factors are not necessarily related to the specific characteristics of the city and
458
TOURISM ECONOMICS
Arch
Museums
Landsc
Atmosph
Business
Culture
Shopping
Business
Nightlife
Entertainment
Motivations
(1)
Arch
Sex
Income
Educ
Satisfaction
Age
Characteristics
Business
(3)
(4)
(4)
Mherit
Tangible
(2)
Monum
Museums
Landsc
Intangible
Holiday
Tradition
Customs
Knowl
National
Loyalty
Return
Recomm
Figure 1. Conceptual model of the tourist–loyalty process as function of a
tourism complex.
the satisfaction achieved with its different elements and services, the analysis
focuses on pull factors. A first important aspect to be analysed is the relationship
between the characteristics of tourists (reason to travel, nationality, age, sex,
level of income, level of education or membership of a heritage club) and their
pull motivations (business, shopping, nightlife, atmosphere, cultural events,
museums, architecture or landscape) to visit a tourism destination (represented
by arrow 1 in Figure 1). Martín Armario (2008) describes this interrelation
between – before journey – personal characteristics and intended behaviour at
the destination (that is, the motivation to travel) as the first phase in tourist
behaviour.
The second phase concerns the experience at the destination (Martín Armario,
2008). The link between the pre-journey motivations and satisfaction is
identified by Lubbe (1998), who stresses that the motivation to travel is linked
with the awareness of certain needs and the realization that a particular
destination might serve to fulfil those needs. According to the motivations with
which specific segments of the tourist market travel, different sets of products
and services might be enjoyed and the satisfaction obtained could differ. This
relationship could allow for a better understanding between the pull aspects of
the destination that are important for specific segments and the satisfaction
provided by these, leading to a better segmentation of specific tourist groups
(Chi and Qu, 2008; Lee, 2009). Therefore, a second aspect to be analysed in
this work is the relationship between the motivations expressed by tourists
visiting a tourism destination and the levels of satisfaction they obtained with
the different aspects of the city (represented by arrow 2 in Figure 1).
Culture, product differentiation and market segmentation
459
While satisfaction with the destination resources is thus directly related to
travel motives, personal characteristics are hypothesized to have both an indirect
effect via the motivational travel aspects, and a direct effect on satisfaction. As
a primary hypothesis, personal characteristics are expected to influence mainly
the motives that are important for visiting Amsterdam, defining different
tourist segments. Next, these motives can be considered as expectations of the
destination, with the tourist deciding on a specific destination because it is seen
as a means to serve the motivational need (Aziz and Ariffin, 2009). If
expectations regarding these motives are met, this should result in satisfaction.
However, some personal characteristics might directly affect the level of
satisfaction received from a tourist experience, irrelevant of the motivation for
travelling. This would suggest that some market segments are more likely to
be (dis)satisfied due to personality traits or other socio-demographic variables
(represented by arrow 3 in Figure 1).
Furthermore, assuming that the loyalty to a tourism destination is related
to the satisfaction obtained on a previous visit or previous visits, as is commonly
assumed in the literature, it is important to understand how the satisfaction
with each aspect of the tourism supply influences the loyalty to the destination.
As the overall satisfaction of a tourist results from the satisfaction obtained from
each of their experiences with different services and elements of the tourism
supply, all these elements contribute to the loyalty of tourists regarding a
destination (Castro et al, 2007; Lee, 2009). This loyalty can be evaluated by
taking into consideration the intention of the tourists to return and/or to
recommend the visit to their families and friends (Oppermann, 2000; Yoon and
Uysal, 2005; Chen and Tsai, 2007). Accordingly, the fourth aspect to be
analysed is the relationship between the satisfaction obtained with the different
aspects of the city and the loyalty to Amsterdam as a tourism destination
(represented by arrow 4 in Figure 1).
The analysis developed in the present study starts from the segmentation of
the tourism market, considering the different characteristics of tourists in order
to identify their motivations, the relationship of these personal characteristics
and motivations to the level of satisfaction obtained with each aspect of the
tourism supply, and the implications of the satisfaction for the loyalty to the
destination. Finally, the direct relationship between the characteristics of the
tourists (segmentation) and the loyalty will be analysed (represented by arrow 5
in Figure 1). The architecture of the conceptual model with the relationships
to be analysed in our study is shown in Figure 1.
Our statistical analysis will be performed using a structural equation model
(SEM). Some recent examples of the use of SEMs are found in Chen and Chen
(2010), Chen and Tsai (2008), Lee et al (2007) Lee and Hsu (2013) Yoon and
Uysal (2005), all of them analysing the relationship between motivations
and loyalty. Castro et al (2007), Chen and Tsai (2007), Chi and Qu (2008) and
Lee (2009) include the concept of ‘image’ of a destination to analyse the
loyalty of tourists. Dyer et al (2007) model the residents’ perceptions related
to tourism development, Abrate et al (2011) use a SEM to analyse the
relations between characteristics of places and hotels, reputation, quality and
prices, while Kim and Li (2009) investigate the relation between customer
satisfaction and loyalty, taking into consideration the transaction costs over the
Internet.
460
TOURISM ECONOMICS
Table 1. Characteristics of the variables used.
Gender (female)
Heritage membership (Yes)
Main travel purpose
Holiday
Business
Other
Age
< 18
18–34
35–54
55–74
> 74
Income
< €15,000
€15,000–25,000
€25,000–35,000
€35,000–45,000
€45,000–55,000
> €55,000
Education
Pre high school
High school
Vocational
Bachelor’s degree
Master’s degree
Nationality
Dutch
Other European
Other
Motivations
Architecture
Museums
Landscape
Cultural events
Shopping
Business
Nightlife
Atmosphere
Appreciation
Architecture
Monuments
Museums
Urban landscape
Cultural events
Traditions
Customs
Knowledge
Measurement
level
Full sample
(n = 645),
frequency/mean (sd)
Missing deleted
(n = 523)
frequency/mean (sd)
Categorical
Categorical
Binary (Yes/No)
52.1%
20.5%
51.4%
22.4%
67.0%
10.5%
22.5%
73.4%
10.7%
15.9%
6.4%
62.8%
22.2%
8.4%
0.3%
5.2%
65.6%
21.8%
7.1%
0.4%
32.7%
18.2%
13.8%
9.5%
5.6%
20.3%
32.9%
18.2%
14.0%
9.6%
5.7%
19.7%
5.3%
22.0%
11.9%
35.0%
25.7%
4.0%
22.0%
11.5%
35.9%
26.6%
20.2%
49.6%
30.2%
19.7%
50.3%
30.0%
48.2%
65.1%
33.2%
21.6%
37.2%
6.5%
39.4%
54.0%
48.9%
63.9%
33.5%
22.0%
37.3%
6.9%
40.2%
55.4%
4.10 (0.922)
3.69 (1.023)
4.03 (0.987)
3.76 (0.967)
3.54 (1.148)
3.83 (1.086)
3.43 (1.162)
3.35 (1.100)
4.09 (0.942)
3.65 (1.047)
4.04 (0.990)
3.78 (0.976)
3.55 (1.151)
3.84 (1.068)
3.44 (1.152)
3.34 (1.112)
Ordinal
Ordinal
Ordinal
Categorical
Binary (Yes/No)
Ordinal
Culture, product differentiation and market segmentation
461
Methodology: a structural equation model
The conversion of the conceptual model in Figure 1 into an operation
measurement model calls for an extensive database. From the multi-purpose
survey questionnaire (1), we concentrated on four aspects specifically: motives
for a visit; loyalty to the destination; appreciation of the cultural heritage
present; and personal characteristics of the visitors. Table 1 provides an overview
of the variables concerned.
The hypothesized relationships between the exogenous and endogenous
variables of the model were tested with the AMOS 19 SEM package for SPSS.
In an initial phase, the data were investigated for missing values. Since a
number of SEM functionalities require a complete data set, it is advisable to
either delete or impute cases which contain missing values. Of the 645 cases,
a total of 122 cases contained at least one missing value, primarily on the
income variable (n = 107). A missing-data pattern analysis did not reveal any
significant association with the scores on the related variables, which indicates
that a simple deletion of cases with missing values would not result in serious
estimation errors. Since data-imputation would always imply the artificial
construction of variable scores, and since the number of missing values to
estimate is comparatively large, our further SEM analysis has only taken
complete cases into account.
To test our hypothesized model as shown in Figure 1, Mulaik and Millsap’s
(2000) four-step modelling approach was used, consisting of:
(a)
(b)
(c)
(d)
explanatory factor analysis to establish the number of latent variables;
confirmatory factor analysis to confirm the measurement model;
a structural model to test the relationships between the model variables;
nested models testing in order to identify the most parsimonious model.
While Steps (b) to (d) are intrinsic to SEM, the first step is performed in
commonly used statistical software packages. The unidimensionality of each
proposed construct, a necessity in the model building step, was assessed by
principal component analysis (PCA) in SPSS 17.0 (Sethi and King, 1994). Since
the variables used in the analysis were on either ordinal or dichotomous levels,
a polychoric and tetrachoric correlation matrix was used instead of the more
common Pearson’s product-moment correlation (Jöreskog and Sörbom, 1996).
Table 2 gives an overview of the PCA results. Explanatory factor analysis on
motivations appears to lead to a three-factor solution, based on the Kaiser
criterion (eigenvalues above 1), with a cumulative explained variance of 65%.
The results are in line with previous theoretical analysis (Manzanek, 2010),
dividing the motivational factor into a cultural, a business and an entertainment
motive. The first identified factor, (1) the cultural motivation, incorporates the
visiting purposes of architecture, museums, urban landscape, cultural events and
the general atmosphere of the city. The second, (2) entertainment, is largely
constructed from the items shopping and nightlife. The third, (3) business, is
mainly concerned with one variable: visits that have business as a driving factor.
The PCA applied to the satisfaction variables yields a two-factor solution
with a cumulative explained variance of 62%. Table 2 gives an overview of the
construction of the two satisfaction factors. Factor 1 combines all variables
462
TOURISM ECONOMICS
Table 2. Varimax rotated component matrix of motivation and satisfaction.
Items
Factors
1
2
3
Motivations
Activities planned architecture
Activities planned museums
Activities planned landscape
Activities planned cultural events
Activities planned shopping
Activities planned business
Activities planned nightlife
Activities planned atmosphere
0.790
0.660
0.789
0.403
–0.026
–0.030
0.282
0.686
0.094
0.284
–0.082
0.272
0.895
0.063
0.652
0.260
0.081
–0.403
0.063
0.560
–0.003
0.839
0.252
0.215
Satisfaction
Appreciation of architecture
Appreciation of monuments
Appreciation of museums
Appreciation of urban landscape
Appreciation of cultural events
Appreciation of traditions
Appreciation of customs
Appreciation of knowledge
0.138
0.160
0.065
0.158
0.455
0.891
0.886
0.850
0.824
0.767
0.727
0.687
0.226
0.068
0.122
0.123
concerning the satisfaction with intangible heritage: traditions, customs and
knowledge, while the second component, Factor 2, depends mainly on
satisfaction with the tangible heritage, namely, the city’s architecture, monuments,
museums and urban landscape. This distinction between the role of tangible
and intangible cultural aspects of the city is the major theoretical contribution
of this work, leading to important practical and managerial consequences.
After conducting this exploratory factor analysis, a confirmatory factor analysis
– to test the adequacy and validity of individual items and latent variables –
and SEM – to test the significance of the hypothesized paths between all
variables – was performed in AMOS. Both maximum likelihood and Bayesian
estimation were used. A Bayesian approach is normally advocated in case of
ordinal or dichotomous measurement levels and a non-normal distribution, both
observed in the data. However, it should be noted that various authors have
observed only marginal differences between maximum likelihood and Bayesian
estimation outcomes (Byrne, 2010). Therefore, the decision was taken to run
a simultaneous maximum likelihood estimation in order to compare the results
and model fit indices.
Results
The model fit indices for the original confirmatory factor measurement model
indicate a model misspecification with a chi-square (χ²) statistic of 664.745
(df = 229, p-value = 0.000), a normal chi-square (χ²/df) of 2.903, a root mean
Culture, product differentiation and market segmentation
463
square error of approximation (RMSEA) of 0.060 and a comparative fit index
(CFI) of 0.890 (for a discussion of thresholds, see Wheaton et al, 1977; Steiger,
2007 and Tabachnick and Fidell, 2007). Apart from looking at the total model
fit indices, the significance of the individual factor loadings should also be
considered (Hooper et al, 2008). None of the measurement items had nonsignificant factor loadings at a 99% confidence interval, while the mean values
and standard errors of the items (as seen in the appendix) show, on average
sufficient scores to use in a SEM. Indeed, all but two standardized regression
weights were above the minimal level of 0.30 (Merenda, 1997; Hair et al, 1998).
Only the item indicating a shopping motivation (0.270) and the item
measuring satisfaction through cultural events (0.295) were below this threshold
value. The Bayesian estimation procedure shows largely comparable parameter
estimates.
The measurement model was improved by correlating measurement errors
between a number of motivational and satisfaction items, since it could be
presumed that the response on travel motive and satisfaction is influenced by
the same underlying structures. Furthermore, both the ‘appreciation of cultural
events’ and the ‘motivation for cultural events’ were deleted from their latent
constructs, since the standardized regression weights were below or only
marginally above the 0.30 level. The shopping motive was kept in the analysis
for its theoretical value. Apart from a χ²-value of 337.607 (df = 177, p-value
= 0.000) – which is known to be sensitive to sample size, departures from
multivariate normality and model complexity (Schumacker and Lomax, 2004)
– all other model fit indices now show an acceptable value (χ²/df = 1.907,
RMSEA = 0.042, CFI = 0.956). Furthermore, analysis of both convergent –
the extent of internal factor consistency – and discriminant – the extent of
distinctness between factors – validity supported the model structure. Convergent
validity can most easily be tested in confirmatory factor analysis by inspecting
the significance of factor loadings of each measurement item (Bagozzi et al,
1991; Hill and Hughes, 2007). All factor item scores of our final measurement
model were found significant on a 99% confidence level, while only the
standardized score of shopping stayed below the 0.30 threshold, as can be
observed from Table 3.
Discriminant validity was tested by assessing whether the square root of the
average variance extracted (AVE) of each latent construct was larger than the
correlation between different latent constructs. Furthermore, the AVE for each
construct should at least be as high as 0.50 (Zait and Bertea, 2011). Table 4
gives an overview of the AVE – on the diagonal – as compared to the interfactor correlations. Since all squared AVE-values are much larger than the
correlations between different factors and the non-squared values are above 0.50,
we can assume that the latent factors are sufficiently dissimilar.
In a next step, causal relationships were added according to the theoretical
model of Figure 1 and tested on path significance. The results of the initial
Maximum Likelihood estimation procedure revealed acceptable model-fit
indices (χ²/df = 2.166, RMSEA = 0.047, CFI = 0.941), even though the
χ²-value = 400.699 (df = 185, p-value = 0.000) was significant. However, only
a limited number of regression paths were found to have significant estimates
at a 95% confidence level. Comparable results were found with the use of
Bayesian estimation. To test the impact of these non-significant relationships,
464
TOURISM ECONOMICS
Table 3. (Un)standardized factor weights of measurement items.
Factor
Item
Unstandardized
estimates (SE)
Standardized
estimate
Motive Culture
Architecture
Museums
Landscape
Atmosphere
1.000
0.986 (0.102)***
0.779 (0.094)***
0.813 (0.101)***
0.574
0.584
0.469
0.467
Motive Shop
Shopping
Nightlife
1.000
3.424 (1.074)***
0.271
0.915
Satisfaction Tangible
Architecture
Monuments
Museums
Urban Landscape
1.000
0.996 (0.088)***
0.760 (0.075)***
0.766 (0.075)***
0.737
0.656
0.529
0.543
Satisfaction Intangible
Traditions
Customs
Knowledge
1.000
1.173 (0.066)***
0.972 (0.060)***
0.785
0.854
0.733
Loyalty
Recommend
Return
1.000
0.921 (0.033)***
0.963
0.841
Note:
*
p-value < 0.05; ** p-value < 0.01, *** p-value < 0.001.
Table 4. Average variance extracted analysis.
Motive culture
Motive shop
Satisfaction tangible
Satisfaction intangible
Loyalty
Motive
culture
Motive
shop
Satisfaction
tangible
Satisfaction
intangible
0.903
0.347
0.205
0.372
0.637
0.990
–0.063
0.266
0.303
0.754
0.318
–0.065
0.841
0.344
Loyalty
0.985
paths were deleted one by one, based on their significance levels and the value
of the standardized regression estimates (with the lowest values deleted first).
A likelihood ratio test could then be performed on each step to investigate
whether the relationship could be erased from the model on grounds of
parsimony without significantly affecting the model fit.
In Table 5 we compare the difference between the χ²-value of the less
parsimonious model with the χ²-value of the nested model with that between
the tabled χ²-values for the related degrees of freedom. While in general a lower
value is preferred, for an insignificant difference the modification should be
accepted on grounds of parsimony. Since the tabled χ²-value for one degree of
freedom (that is, one extra path deleted) and an α-level of 0.05 is 3.48, we can
conclude that only the final step in Table 5 shows a significant difference in
χ²-value, thus giving preference to the more constricted model on grounds of
Culture, product differentiation and market segmentation
δ1
δ2
δ3
δ4
δ5
δ6
δ7
Arch
Museums
Landsc
Atmosph
Business
Shopping
Nightlife
M1
ε1
ε2
M2
M3
Holiday
465
ε3
ε4
Business
Arch
δ8
Monum
δ9
S1
Age
Museums
δ10
Sex
Landsc
δ11
Income
ε5
Educ
S2
Mherit
Dutch
L
Return
δ15
Tradition
δ12
Customs
δ13
Knowl
δ14
ε6
Recomm
δ16
Figure 2. Operational structural equation model.
Notes: For reasons of readability, covariances are not shown in the model. Only significant paths are
shown. M1 = culture motive; M2 = business motive; M3 = entertainment motive; S1 = satisfaction
tangible; S2 = satisfaction intangible; L = loyalty, χ² (204) = 390.872, p = 0.000, χ²/df (1.916),
RMSEA = 0.043, CFI = 0.942.
parsimony in all other 36 cases. Furthermore, the variable ‘other European’
could be deleted altogether since all regression weights were set to zero, further
facilitating the model.
In the final structural model, 27 regressions were included in the model, 26
of which were significant on an α-level of 0.05 or lower – only the t-test of
the relationship between business and loyalty had a p-value slightly above 0.05
(p = 0.056). The overall model fit statistics indicate that the accepted model
fits the data better than the original model. While we have to acknowledge
that the χ²-value of our final model remained significant, it is a known problem
that these indices are biased with small sample sizes, a large number of
variables, and a non-normal data distribution (Kenny and McCoach, 2003;
Schumacker and Lomax, 2004; Fan et al, 2011). Since the better performing
goodness of fit indices (root mean square error of approximation, χ²/df, and
comparative fit index) indicate a reasonable to good model fit for the final
model, the final parameter estimates can be considered as sufficiently stable.
Table 6 gives an overview of the significant unstandardized regression weight
estimates with both maximum likelihood method (left side of Table 6) and
Bayesian estimation (right side of Table 6). Note that since Bayesian estimates
are based on a 95% confidence interval around the mean, the level of significance
for these estimates is maximized at an α-level of 0.05. Both estimates under
maximum likelihood and Bayesian estimation generate comparable results, with
consistent unstandardized factor weights and significance. However, as indicated
466
TOURISM ECONOMICS
Table 5. Likelihood ratio analysis of nested models.
Steps
Path deleted
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Full model
Other European → L
Other European → M2
Holiday → L
Holiday → M2
Other European → M3
Education → M1
Other European → M1
Age → M2
Business → S1
Heritage membership → S2
Gender → M1
Heritage membership → M2
Holiday → S2
Education → M2
Heritage membership → M3
Income → M1
Business → S2
Gender → M2
Age → M1
M2 → S1
Income → M3
Income → S2
Age → S2
Heritage membership → S1
Education → S2
Business → M1
Other European → S2
M3 → S1
Business → M3
Other European → S1
Dutch → S2
Gender → L
Income → S1
Age → L
Dutch → M2
Business → L
df
⎟²
D²
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
400.699
400.699
400.701
400.707
400.716
406.707
406.826
406.966
407.038
409.150
409.206
409.276
409.370
409.474
409.674
409.842
410.064
410.387
410.796
411.255
411.747
412.544
413.434
413.898
414.695
415.547
416.413
417.452
418.616
419.859
421.133
422.126
424.032
426.245
428.570
431.089
434.709
0.000
0.002
0.006
0.009
5.991
0.119
0.140
0.072
2.112
0.056
0.070
0.094
0.104
0.200
0.168
0.222
0.323
0.409
0.459
0.492
0.797
0.890
0.464
0.797
0.852
0.866
1.039
1.164
1.243
1.274
0.993
1.906
2.213
2.325
2.519
3.620
Note: M1 = culture motive; M2 = business motive; M3 = entertainment motive; S1 = satisfaction
tangible; S2 = satisfaction intangible; L = loyalty.
by other authors (Nevitt and Hancock, 2001; Mîndrilã, 2010), the standard
errors of the maximum likelihood estimates are inflated under violations of
multivariate normality and inadequate measurement level.
As it could be expected, results from the econometric analysis confirm that
a holiday purpose for travelling has a significant positive influence on both the
cultural and entertainment motive, while, conversely, the business purpose has
a significant positive correlation with the business motive. Another factor
Culture, product differentiation and market segmentation
467
Table 6. Unstandardized path estimates with maximum likelihood and Bayesian
estimation.
Maximum likelihood,
unstandardized estimates
(SE)
Motive Culture
Nature holiday
Heritage membership
Dutch nationality
Bayesian estimation,
unstandardized estimates
(SE)
0.139 (0.031)***
0.064 (0.030)*
–0.472 (0.045)***
0.137 (0.001)*
0.064 (0.001)*
–0.467 (0.001)*
0.295 (0.033)***
0.017 (0.005)**
0.295 (0.001)*
0.017 (0.000)*
Motive Entertainment
Nature holiday
Age
Sex
Educational level
Dutch nationality
0.027 (0.012)*
–0.036 (0.010)***
–0.041 (0.012)***
–0.009 (0.004)*
–0.042 (0.015)**
0.030 (0.000)*
–0.041 (0.000)*
–0.047 (0.000)*
–0.010 (0.000)*
–0.047 (0.000)*
Satisfaction tangible
Nature holiday
Age
Sex
Education level
Dutch nationality
Motive culture
–0.351 (0.093)***
0.129 (0.051)*
0.137 (0.067)*
0.119 (0.028)***
0.579 (0.176)***
1.578 (0.331)***
–0.354 (0.002)*
0.129 (0.001)*
0.137 (0.001)*
0.118 (0.000)*
0.582 (0.005)*
1.607 (0.010)*
Satisfaction intangible
Sex
Motive culture
Motive business
Motive entertainment
0.210 (0.074)**
1.257 (0.185)***
0.314 (0.144)*
1.135 (0.372)**
0.212 (0.002)*
1.274 (0.005)*
0.312 (0.003)*
1.050 (0.011)*
Loyalty
Satisfaction tangible
Satisfaction intangible
Nature business
Income
Education level
Heritage membership
Dutch nationality
–0.051 (0.016)**
0.043 (0.013)**
–0.060 (0.031)
–0.018 (0.005)***
0.016 (0.008)*
–0.052 (0.023)*
–0.688 (0.026)***
–0.051 (0.000)*
0.043 (0.000)*
–0.060 (0.001)
–0.018 (0.000)*
0.016 (0.000)*
–0.052 (0.001)*
–0.689 (0.001)*
Motive Business
Nature business
Income
Note: * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001.
positively influencing the business travel motivation is the income variable with
higher-income categories more likely to visit Amsterdam for business purposes.
This is also an expected result but, interestingly, the results also indicate that
tourists in higher age categories seem less motivated by the entertainment
opportunities in Amsterdam. This might be caused mainly by the presence of
‘shopping’ in this factor, which requires a certain income. Both the educational
level and sex of the respondent have a significant negative effect on entertainment
468
TOURISM ECONOMICS
as a travel motive, implying that men and higher educated tourists are less
likely to travel to the destination for shopping and nightlife than women and
lower-educated visitors. Furthermore, the entertainment motive seems less
significant for Dutch tourists than for non-Dutch visitors. Finally, heritage
membership has a significant positive influence on the cultural motive and this
motivation is less important for Dutch nationals as compared with other
nationalities, as it could be expected.
Four of the six possible paths between motivations and satisfaction with the
urban resources were found significant on an α = 0.05 level. The statistical
analysis indicates the existence of a highly positive relationship between the
cultural motive and the satisfaction with both tangible and intangible heritage
in Amsterdam. Business travellers and tourists with an entertainment motive
also appear to record comparatively higher satisfaction with intangible heritage:
the traditions, customs and local knowledge at the destination. The positive
relationship between sex and both tangible and intangible satisfaction is indicative
of the fact that men seem to be more satisfied with Amsterdam’s tangible and
intangible heritage. Other personal characteristics that further advance satisfaction
with tangible city elements are age, education level and Dutch nationality.
These factors all increase the chance of tourists being satisfied with the built
environment. Finally, tourists travelling primarily with a holiday motive are
generally less satisfied with tangible aspects. These results show the importance
of the distinction between tangible and intangible heritage, since different
variables can have diverse impacts on both types of tourist satisfaction.
Next, from Table 6 we can deduce that both latent satisfaction variables can
be significantly related to loyalty. However, while a higher satisfaction with
intangible heritage leads to a higher loyalty (as it was expected), satisfaction
with the tangible heritage on the destination appears to have a negative impact
(which is an important and innovative result of this study). These different
impacts of satisfaction with tangible and intangible cultural assets on the
loyalty of tourists form a second major result that arises from the distinction
between different cultural aspects of the city. Furthermore, four relationships
between loyalty and personal characteristics had significant regression weights.
Income has a negative weight, indicating that higher income categories show
less loyalty in our dataset. Other negative relationships are found for heritage
membership and Dutch nationality, while education has a significant positive
value.
A final point of interest is the possible indirect effects between the different
travel motives and loyalty. While we did not hypothesize a direct relationship
between these variables, significant indirect effects can still be estimated due
to the paths connecting motives, significance and loyalty. The results of this
analysis indicate a positive indirect relation between the entertainment (0.049)
and business (0.013) motive with loyalty, while cultural travel motives (–0.026)
have a negative indirect effect.
Discussion
A number of significant relationships found are intuitively meaningful and
show the hypothesized signs. As expected, visitors coming to Amsterdam for
Culture, product differentiation and market segmentation
469
business purposes are motivated more by business opportunities in Amsterdam,
while holiday purposes can be related positively to the entertainment and
cultural motives. Another interesting aspect of the business motive is its
correlation with a comparatively higher income, empirically validating the
(expected) higher spending power of business travellers and making these
visitors an interesting marketing group to target. Furthermore, travellers with
a primarily business motive reported higher satisfaction with intangible
heritage, indirectly leading to a significant positive effect on the return and
recommendation potential of this tourist group. The fact that business travellers
show higher rates of satisfaction for intangible heritage might be related to both
their location in the city – with business districts often located outside of
historic centres – and the time of day when they have leisure time. This group
might therefore have a higher chance of experiencing intangible heritage than
physical, tangible relicts. Furthermore, their business trip could trigger a future
leisure return visit if the city is found enjoyable, therefore potentially increasing
loyalty.
While, as already mentioned, the cultural motive was related positively to
the holiday travel purpose, this travel motive was furthermore of primary
importance for international tourists and members of a heritage organization.
Both results are expected and acknowledge the image of Amsterdam as an
international cultural destination. Nevertheless, tourists visiting the city with
a holiday motive achieve less satisfaction with tangible elements, which reveals
the importance of the intangible aspects of local culture on the satisfaction of
visitors on holidays in Amsterdam. The cultural aspects seem of less importance
to Dutch visitors, who are more likely to travel for alternative purposes (such
as visiting friends or relatives or for administrative reasons). Additionally, these
national tourists reported less interest in shopping and nightlife as their main
travel reason. This travel motive could further be related to younger people,
females and people with a lower education, thereby possibly uncovering another
aspect of the Amsterdam image – the libertarian atmosphere related to its drugs
and prostitution laws – while also showing the importance of its shopping
districts to a certain segment of travellers.
While business travellers were, in general, more satisfied with the intangible
aspects of Amsterdam, culturally motivated tourists indicated a higher
satisfaction on both the intangible and the built environment: the architecture,
museums, monuments and urban landscape. This result is sensible, since this
subgroup of visitors attaches primary importance to these aspects of the city.
The estimated positive relationship seems to imply that Amsterdam lives up
to the expectations of cultural tourists. At the same time, shopping and
nightlife tourists seemed to experience low levels of satisfaction with the
tangible elements, possibly as a result of a lack of interest in these destination
aspects. Furthermore, it might be argued that the historical character of the
inner city, with its generally smaller streets, affects the shopping experience
since the main shopping street in the centre of Amsterdam: Kalverstraat, is
sensitive to crowding issues.
An important and perhaps counterintuitive empirical observation relates to
the association between satisfaction and tourist loyalty. Specifically, tourists who
achieve higher levels of satisfaction with intangible aspects of the city tend to
be more loyal while the opposite holds true in case of satisfaction with tangible
470
TOURISM ECONOMICS
heritage. An explanation might be that tourists who mainly experience
satisfaction with the built environment have a more shallow relationship with
the destination. This segment of cultural visitors might then comprise of
‘collectors’ of cultural experiences; travellers who visit a destination once to be
able to check it from the list and then move on to another cultural attraction.
As a result, these visitors, limiting themselves to visiting city areas, museums
and landmark attractions, could be considered less loyal towards the destination,
explicitly considering the return potential. Conversely, travellers who are both
motivated by and satisfied with intangible cultural elements might have a
greater potential for return visits since an intangible experience of the local
culture could be more difficult to achieve in a single visit. Furthermore,
intangible aspects have a deeper effect, relating to lifestyles instead of aesthetics.
Both relationships seem further acknowledged by the significant indirect
relationships between motivations and loyalty, where estimating the indirect
effects led to the detection of a significant positive relationship between the
entertainment and the business motive, while cultural travellers were found to
be less loyal towards the specific destination.
Finally, a number of personal tourist characteristics were found to be directly
related to loyalty to Amsterdam. The lower loyalty of heritage conservation
organization members confirms our results concerning the cultural motive. A
significantly positive effect on loyalty was found for tourists in the lower income
categories, who might be more limited in their travel possibilities, tourists with
a higher level of education and non-European tourists.
Conclusion
The primary goal of this study was to assess the impact of several factors on
the loyalty (measured by the possibility of a return visit or a recommendation
to visit) to the destination of Amsterdam, since this loyalty can offer a
competitive advantage to the destination in attracting future tourists. The most
interesting conclusion to note, concerning the effect of different factors on
visitor loyalty is the observation that higher satisfaction does not necessarily
increase tourist retention.
The distinction between the role of tangible and intangible cultural assets
– and its implications on the satisfaction and loyalty observed for different
groups of tourists – is the most interesting theoretical and practical contribution
of this analysis. In fact, it was observed that intangible sources (such as
traditions and customs) have a positive effect on the chance for a positive
recommendation or a return visit; however, it is noteworthy that a high
satisfaction with tangible resources (such as museums, architecture and urban
landscape) had the opposite effect. While this seems to conflict with theoretical
expectations and the popular belief, it might be explained by the fact that the
group of tourists looking for these specific experiences is primarily focused on
once in a lifetime experiences and will move on to other destinations once these
needs are met.
This has some relevant implications for tourist marketing strategies in
Amsterdam, revealing that the preservation of the local life style and traditions
plays an extremely important role regarding the satisfaction and loyalty of
Culture, product differentiation and market segmentation
471
tourists. Thus, it is important to take into consideration that the construction
of attractions oriented for tourists can have negative consequences on the daily
life of the local community, also with undesirable results regarding the
satisfaction of tourists with these intangible elements of the city.
Nonetheless, the city has shifted its attention to the higher spending market
of cultural tourists, in place of backpackers and party tourism. However, such
a strategy might have unfavourable effects in the longer run, since the latter
groups are more likely to return in the future (possibly as cultural tourists at
a later stage in their lives) while the former group, seeking a purely cultural
experience, is less loyal towards the city. It might thus be advisable not to limit
marketing efforts to a single segment of the tourist market and also to focus
on the lifetime value of a tourist – especially considering the fact that higher
income groups are also less likely to return.
While the results show an interesting, somewhat unexpected relationship
between satisfaction and loyalty, studies in other destinations are needed to
validate the results in a larger context. The city of Amsterdam is a very
particular tourist destination, with part of its international attraction due to
its liberal image concerning gay rights, squatting, prostitution and soft drugs.
As a result, the importance of the intangible heritage on tourist loyalty may
be quite specific to the nature of the destination and the results may not be
easily reproducible in other situations. Since the European framework that
provided the database collected data in a number of destinations, this opens
up possibilities to attempt to fit a similar model in other cities.
Endnotes
1. The data used for the empirical part of this paper were mostly collected within the Sixth
Framework Programme from European Union (FP6 EU) project, ‘Integrated e-Services for
Advanced Access to Heritage in Cultural Tourist Destinations’ (ISAAC). The data were collected
by user surveys carried out in the city of Amsterdam between August and November 2007 as
part of a broader multi-purpose tourist investigation. These surveys involved extensive field data
collection by interview teams hired and trained by the University of Nottingham, one of the
ISAAC partners. Three different groups of people were targeted: residents, visitors (tourists) and
service providers in the tourist sector. The questionnaires used both online and face-to-face
interview mode (stand-alone computer versions or paper versions). In total, 31% of responses
were made online, using the ISAAC website survey; 24% were done on a computer version using
a laptop; and 45% were done on paper (see also ISAAC D1.4, 2007). The fieldwork developed
by VU Amsterdam was exclusively focused on tourists visiting Amsterdam, where approximately
650 tourists filled out a questionnaire.
References
Abrate, G., Caproello, A., and Fraquelli, G. (2011), ‘When quality signals talk: evidence from the
Turin hotel industry’, Tourism Management, Vol 32, pp 912–921.
ATCB (2012), Amsterdam Visitor Survey, Amsterdam Tourism and Convention Bureau, Amsterdam.
Aziz, N.A., and Ariffin, A.A. (2009), ‘Identifying the relationship between travel motivation and
lifestyles among Malaysian pleasure tourists and its marketing implications’, International Journal
of Marketing Studies, Vol 1, No 2, pp 96–106.
Bagozzi, R.P., Yi, Y., and Phillips, L.W. (1991), ‘Assessing construct validity in organizational
research’, Administrative Science Quarterly, Vol 36, No 3, pp 421–458.
Bansal, H.S., and Eiselt, H. (2004), ‘Exploratory research of tourist motivations and planning’,
Tourism Management, Vol 25, pp 387–396.
Buhalis, D. (2000), ‘Marketing the competitive destination of the future’, Tourism Management,
Vol 21, pp 97–116.
472
TOURISM ECONOMICS
Byrne, B.M. (2010), Structural Equation Modeling with AMOS: Basic Concepts, Applications, and Programming, 2nd edn, Routledge, New York, NY.
Castro, C., Armario, E., and Ruiz, D. (2007), ‘The influence of market heterogeneity on the
relationship between a destination’s image and tourists’ future behavior’, Tourism Management,
Vol 28, pp 175–187.
Chen, C., and Chen, F. (2010), ‘Experience quality, perceived value, satisfaction and behavioral
intentions for heritage tourists’, Tourism Management, Vol 31, pp 29–35.
Chen, C., and Tsai, D. (2007), ‘How destination image and evaluative factors affect behavioral
intentions?’, Tourism Management, Vol 28, pp 1115–1122.
Chen, C., and Tsai, D. (2008), ‘Perceived value, satisfaction, and loyalty of TV travel product
shopping: involvement as a moderator’, Tourism Management, Vol 29, pp 1166–1171.
Chi, C., and Qu, H. (2008), ‘Examining the structural relationships of destination image, tourist
satisfaction and destination loyalty: an integrated approach’, Tourism Management, Vol 29, pp 624–
636.
Crompton, J. (1979), ‘Motivations for pleasure vacations’, Annals of Tourism Research, Vol VI, No 4,
pp 408–424.
Dann, G. (1981), ‘Tourist motivation: an appraisal’, Annals of Tourism Research, Vol VIII, No 2,
pp 187–219.
Dyer, P., Gursoy, D., Sharma, B., and Carter, J. (2007), ‘Structural modeling of resident perceptions
of tourism and associated development on the Sunshine Coast, Australia’, Tourism Management,
Vol 28, pp 409–422.
Fan, X., Thompson, B., and Wang, L. (2011), ‘Effects of sample size, estimation methods, and model
specification on structural equation modeling fit indexes’, Structural Equation Modeling:
a Multidisciplinary Journal, Vol 6, No 1, pp 56–83.
Hair, J.F., Anderson, R.E., Tatham, R.L., and Black, W.C. (1998), Multivariate Data Analysis,
Prentice Hall, Upper Saddle River, NJ.
Hill, C.R., and Hughes, J.N. (2007), ‘An examination of the convergent and discriminant validity
of the strengths and difficulties questionnaire’, School Psychology Quarterly, Vol 22, No 3,
pp 1–23.
Hooper, D., Coughlan, J., and Mullen, M.R. (2008), ‘Structural equation modelling: guidelines for
determining model fit’, Electronic Journal of Business Research Methods, Vol 6, No 1, pp 53–60.
ISAAC D1.4 (2007), ‘Report on users requirements for ISAAC e-services using conjoint analysis’,
ISAAC deliverable 1.4, (http://www.isaac-project.eu/publications.asp).
Jöreskog, K.G., and Sörbom, D. (1996), LISREL 8: User’s Reference Guide, Scientific Software,
Chicago, IL.
Kenny, D.A., and McCoach, B.D. (2003), ‘Effect of the number of variables on measures of fit in
structural equation modeling’, Structural Equation Modeling: a Multidisciplinary Journal, Vol 10,
No 3, pp 333–351.
Kim, Y.G., and; Li, G. (2009), ‘Customer satisfaction with and loyalty towards online travel
products: a transaction cost economics perspective’, Tourism Economics, Vol 15, No 4, pp 825–
846.
Kozak, M. (2002), ‘Comparative analysis of tourist motivations by nationality and destinations’,
Tourism Management, Vol 23, pp 221–232.
Kozak, M., and Rimmington, M. (2000), ‘Tourist satisfaction with Mallorca, Spain, as an off-season
holiday destination’, Journal of Travel Research, Vol 38, pp 260–269.
Lee, C., Yoon, Y., and Lee, S. (2007), ‘Investigating the relationships among perceived value,
satisfaction, and recommendations: the case of the Korean DMZ’, Tourism Management, Vol 28,
pp 204–214.
Lee, T. (2009), ‘A structural model to examine how destination image, attitude, and motivation affect
the future behavior of tourists’, Leisure Sciences, Vol 31, pp 215–236.
Lee, T., and Hsu, F. (2013), ‘Examining how attending motivation and satisfaction affects the loyalty
for attendees at aboriginal festivals’, International Journal of Tourism Research, Vol 15, No 1,
pp 18–34.
Lubbe, B. (1998), ‘Primary image as a dimension of destination image: an empirical assessment’,
Journal of Travel and Tourism Marketing, Vol 7, No 4, pp 21–43.
Manzanek, J. (2010) ‘Managing the heterogeneity of city tourists’, in Manzanek, J., and Wöber, K.,
eds, Analysing International City Tourism, 2nd edn, Springler-Verlag, Vienna.
Martín Armario, E. (2008), ‘Tourist satisfaction: an analysis of its antecedents’, Asociación Española
de Dirección y Economía de la Empresa, Vol 17, pp 367–382.
Culture, product differentiation and market segmentation
473
Matias, A., Nijkamp, P., and Neto, P. (2007), Advances in Modern Tourism Research, Springer-Verlag,
New York, NY.
Merenda, P.F. (1997), ‘A guide to the proper use of factor analysis in the conduct and reporting
of research: pitfalls to avoid’, Measurement and Evaluation in Counseling and Development, Vol 30,
pp 156–164.
Mîndrilã, D. (2010), ‘Maximum likelihood (ML) and diagonally weighted least squares (DWLS)
estimation procedures: a comparison of estimation bias with ordinal and multivariate non-normal
data’, International Journal of Digital Society, Vol 1, No 1, pp 60–66.
Mulaik, S.A., and Millsap, R.E. (2000), ‘Doing the four-step right’, Structural Equation Modeling,
Vol 7, pp 36–73.
Neuts, B., Romão, J., Nijkamp, P., and van Leeuwen, E. (2013a), ‘Digital destinations in the tourist
sector: a path model for the impact of e-services on tourist expenditures in Amsterdam’, Letters
in Spatial and Resource Analysis, Vol 6, No 2, pp 71–80.
Neuts, B., Romão, J., Nijkamp, P., and van Leeuwen, E. (2013b), ‘Describing the relationships
between tourist satisfaction and destination loyalty in a segmented and digitalized market’,
Tourism Economics, Vol 19, No 5, pp 987–1004
Nevitt, J., and Hancock, G.R. (2001), ‘Performance of bootstrapping approaches to model test
statistics and parameter standard error estimation in structural equation modeling’, Structural
Equation Modeling: a Multidisciplinary Journal, Vol 8, No 3, pp 353–377.
Oppermann, M. (2000). ‘Tourism destination loyalty’, Journal of Travel Research, Vol 39, pp 78–84.
Schumacker, R.E., and Lomax, R.G. (2004), A Beginner’s Guide To Structural Equation Modelling, Second
Edition, Lawrence Erlbaum Associates, Mahwah, NJ.
Sethi, V., and King, W.R. (1994), ‘Development of measures to assess the extent to which an
information technology application provides competitive advantage’, Management Science, Vol 40,
No 12, pp 1601–1626.
Steiger, J.H. (2007), ‘Understanding the limitations of global fit assessment in structural equation
modeling’, Personality and Individual Differences, Vol 42, No 5, pp 893–898.
Tabachnick, B.G., and Fidell, L.S. (2007), Using Multivariate Statistics, 5th edn, Allyn and Bacon,
New York, NY.
United Nations World Tourism Organization (UNWTO) (2013), ‘International tourism to continue
robust growth in 2013’, press release 13006, Madrid.
Wheaton, B., Muthen, B., Alwin, D.F., and Summers, G. (1977), ‘Assessing reliability and stability
in panel models’, Sociological Methodology, Vol 8, No 1, pp 84–136.
Yoon, Y., and Uysal, M. (2005), ‘An examination of the effects of motivation and satisfaction on
destination loyalty: a structural model’, Tourism Management, Vol 26, pp 45–56.
Zait, A., and Bertea, P.E. (2011), Methods for testing discriminant validity’, Management &
Marketing, Vol 11, No 2, pp 217–224.
474
TOURISM ECONOMICS
Appendix
Statistical Test Results
Table A1. Mean scores and standard errors of latent variable items.
Mean
Architecture
Museums
Landscape
Cultural events
Atmosphere
Shopping
Nightlife
Architecture
Monuments
Museums
Urban_landscape
Cultural_events
Traditions
Customs
Knowledge
Recommend
Return
← M1
← M1
← M1
← M1
← M1
← M3
← M3
← S1
←S1
←S1
←S1
←S2
←S2
←S2
←S2
←L
←L
1.000
0.872
0.696
0.409
0.768
1.000
3.446
1.000
0.976
0.763
0.757
1.000
2.499
2.859
2.406
1.000
0.921
SE
0.090
0.084
0.070
0.090
1.079
0.084
0.076
0.075
0.396
0.451
0.385
0.033
Table A2. Mean scores and standard errors of other exogenous variables.
Nature business
Age
Nature holiday
Gender
Income
Education
Heritage membership
Dutch nationality
Other_European
Business motive
View publication stats
Mean
SE
0.107
1.319
0.734
0.514
1.962
2.591
0.224
0.197
0.503
0.069
0.014
.030
0.019
0.022
0.083
0.053
0.018
0.017
0.022
0.011