Professional Documents
Culture Documents
D
espite the increasing importance of the image generators. Clearly, the brochure is not
Internet as a communication and distri- an extinct product.
bution instrument, brochures continue Despite the importance of tourist brochures as
communication tools, little is known about their
*Correspondence to: Rubn Huertas-Garca, Department effectiveness. Some researchers call attention to
of Economics and Business Administration, Barcelona
University, Avda. Diagonal, 696, Barcelona, Spain. large amounts of money spent by tourism distri-
E-mail: rhuertas@ub.edu buters in brochure production but rarely
Copyright 2012 John Wiley & Sons, Ltd.
R. Huertas-Garcia, M. Laguna Garca and C. Consolacin
evaluation of product or service options offered and Urban 1986, Zeithaml 1988, Dodds et al.,
by various companies and examination of 1991, Ostrom and Iacobucci 1995, Teas and
product attributes such as price and brand. Agarwal 2000, Gallarza and Gil, 2006).
(Gupta et al., 2004). Zhou (1997) evaluated Price is often a conictive element because it
the effectiveness of destination brochures plays a double role as an extrinsic product
requested by travellers, suggesting that of quality cue and a product choice monetary
those who read the brochure, 50% visited the constraint (Erickson and Johansson, 1985). These
destination and most consulted the brochure price roles can be labelled as informational and
during the visit. Andereck (2005) found that allocation roles (Rao and Sattler, 2007). Although
tourists who have no knowledge about a tour- the two roles of price are clear and derive from
ist destination demonstrate greater interest in disparate theoretical conceptions, measure-
information contained in brochures than those ments are difcult and tend to confuse the
who visited the destination previously. two when measured directly. In tourism
This study focuses on the evaluation of alter- literature, Naylor and Frank (2001) suggested
natives, the third stage of the buying process. It that price bundling increases perceptions of
is important to recognize that there exist vari- value for rst-time holidaymakers. A study by
ous market segments with differing tastes and Rewtrakunphaiboon and Oppewal (2008) exam-
preferences; experiments require working with ines whether package information inuences
groups that are as homogeneous as possible so intentions to visit and choice of beach holiday
differences in responses demonstrate combina- destinations. On the basis of extant research,
tions of factors rather than socio-demographic the authors assumed that the importance of an
respondent characteristics (Haaijer and Wedel, attribute increases when products display
2007). In this case, the target market chosen according to the attribute (Rewtrakunphaiboon
was students in their nal year of university and Oppewal, 2008, p.129). Accordingly, nd-
who wished to take a graduation trip for 4 days ings suggest that presenting price as a package
during the Easter holiday. The decision about heading (i.e. presenting packages with only
destination and duration, 4 days prior to Easter price) increases intentions to visit and choice of
day, was static, and participants searched for beach holiday destinations among students.
accommodations in travel agency brochures Services such as travel products are perceived
and on the Internet. as riskier purchases than goods (Zeithaml, 1988).
Because services are more difcult to evaluate
PERCEIVED VALUE AND CONSUMER and riskier to purchase, consumers use diverse
DECISIONS processes and cues for evaluation. Searching
for information is one heuristic used to reduce
Choosing a hotel room or a trip from a brochure risks and help travellers make decisions. Travel-
depends on the perceived utility people expect lers who have little information about a destin-
from the room or trip and the price. The market- ation from internal sources (i.e. memory,
ing literature denes the relationship between friends and relatives) use external sources such
perceived utility and price as perceived value. as brochures and the Internet (Andereck, 2005).
Zeithaml (1988, p.14) proposed a denition: per- For purchasing services, extrinsic attributes
ceived value is the consumers overall assessment become cues when information on intrinsic attri-
of the utility of a product based on perceptions of butes is unavailable (Zeithaml 1988). Several
what is received and what is given. What is researchers develop and test models of percep-
received and what is given represent trade-offs tions of value with particular emphasis on a
between give and get components. Benet buyers extrinsic cues such as price and brand
components include intrinsic and extrinsic attri- name as indicators of quality and value. Dodds
butes, perceived quality and other relevant et al. (1991) conducted a study to examine direct
high-level abstractions, and sacrice components and indirect relationships between price, brand
include price and other non-monetary costs. name and store name, and perceived quality,
Managers and academicians recognize a strong product value and willingness to buy. They
inuence of perceived value on consumer behav- argue that price alone is the most important
iour, making it a priority for researchers (Hauser cue of perceived quality; adding store and brand
Copyright 2012 John Wiley & Sons, Ltd. Int. J. Tourism Res. (2012)
DOI: 10.1002/jtr
R. Huertas-Garcia, M. Laguna Garca and C. Consolacin
name information has moderate effects. If a price dominant criterion for evaluating proles, esti-
changes from low to high, perceived value mated by aggregate multi-nomial logit models
increases and at a point decreases signicantly, (Louviere et al., 2000; Street and Burgess, 2007).
impacting willingness to buy negatively. The Implicit models are compensatory and decom-
negative effect of price on value perceptions positional, enabling analyses of the importance
reduces if brand or store name is provided. Price of each products characteristics, starting with a
has a positive inuence on perceived quality but customers stated preferences. In this study, we
negative effects on perceived value and willing- demonstrate the process of designing a choice-
ness to buy. based conjoint analysis and how it determines
Perceived value also depends on the frame- the value of extrinsic cues that, for a market seg-
work under which customers conduct evalua- ment, are used to choose a hotel room from a
tions (Hempel and Daniel, 1993), dened as a tourist brochure. We use one of the least com-
dynamic variable. Perceptions of value change mon models, ranking, based on a one-sixth frac-
if an assessment is made before or after purchase tion from a 28 full factorial design in eight two-
(Woodruff 1997) and whether it is assessed size blocks in which the second prole of each
before sale, at the moment of sale, at the moment block is a mirror image of the rst.
of use or after use (Moliner et al., 2005). Some
authors consider other factors that alter
Choice-based conjoint analysis
perceived value. Lee et al. (2007) reported that
value depends on a consumers characteristics The process of evaluating the choice of accom-
and the type of product under consideration. modation from a brochure operates within
Value is a latent construct not observed directly discrete choice models (CMs). CM is a family
(Teichert and Shehu, 2007); most studies in this of survey-based methodologies for modelling
area use scales to estimate the latent component. preferences for goods in which goods are
We use stated preferences to estimate the value described by their attribute levels. Respondents
or utility generated by varying proles in the are presented with alternative descriptions of
pre-purchase stage. products differentiated by attributes and levels
and rank the various alternatives, rating them
or choosing those they prefer (Hanley et al.,
METHODOLOGY 2001). The form of evaluation (e.g. lexicographic,
elimination by aspect, economic and attribute
Because perceived value and utility are unob- screening) that predominates among respon-
servable, they require scales or proxy variables dents in driving such selections remains elusive
to estimate them (Teichert and Shehu, 2007). (Scarpa and Rose, 2008).
According to Klein (1990), there are two meth- Choice model based on choosing the best
ods to identify and measure consumer needs option has several advantages over traditional
and desires: asking directly or deducing the ratings and rankings. Choice is often the behav-
motives from other kinds of data. The advan- iour of ultimate interest in a decision process
tages of indirect methods include establishing because it estimates behaviours accurately.
robust causal relationships between attributes Models estimated from choice allow direct pre-
and customer evaluations. A popular device, diction of choices, avoiding the need for conjoint
choice-based conjoint analysis, is an indirect simulators (Elrod et al., 1992). Its primary disad-
method used to obtain measurement of prefer- vantage is that it works poorly with small
ences (Haaijer and Wedel, 2007). During conjoint samples because the process of choice contains
analysis evolution over the last 40 years, there minimal information concerning consumer
was a shift in the types of responses used for preferences. A chosen option indicates which
analysis from ranking, to ratings, and nally to alternative is the most preferred, but it does not
choosing the best proles (Elrod and Chrzan, provide information about other alternatives
2007). The most frequent criterion for assessing not selected. A choice process implicitly entails
a choice set was ratings of full proles, evaluated consideration of multiple alternatives, but only
individually (Wittink and Cattin, 1989). Since one option is chosen; in a ranking process, all
the late 1980s, the choice prevailed as the alternatives are considered (Elrod and Chrzan,
Copyright 2012 John Wiley & Sons, Ltd. Int. J. Tourism Res. (2012)
DOI: 10.1002/jtr
Understanding Tourist Decisions
2007). We propose a ranking model adjusted and tendency to rent a room. The theory of
by a proportional odds ordinal logistic model reasoned action posits behaviour intentions
(Train, 2009). The decision to use this model inuences behaviours (Ajzen and Madden,
was based on two objectives. Estimations made 1986). Social psychology research suggests
by using all ranking information are more ef- intentions are the best predictor of behaviours
cient, and efciency is greatest when a full pro- because they allow individuals to incorporate
le design is used. This improvement in all relevant factors independently that inu-
efciency allows more acceptable results with ence behaviours (Fishbein and Ajzen, 1975).
reduced samples. Ranking is consistent with Several studies examine the relationship
consumer behaviour in both economy theory between purchase intentions and purchase
and with respect to the nature of preferences behaviours for non-durable goods (Dowling
representing an ordered relationship (Frank, and Staelin, 1994). The degree to which people
2009). A typical choice-based conjoint analysis express preferences is a reasonable predictor of
is characterized by several stages: purchase behaviours.
Identication of relevant attributes of goods Qualitative research. Investigating the rst two
to be valued. The use of literature reviews, stages of the experimental process of conjoint
consultation with experts and qualitative selection, we began with qualitative research
techniques such as focus groups establishes to identify the attributes considered relevant
these attributes. in the perceived value of choosing a hotel
Assignment of a feasible, realistic range of displayed in a travel agents brochure. We evalu-
plausible levels for attributes. Qualitative ated aggregate utility generated by the sample of
research or literature searches establish ap- students from varying combinations of attri-
propriate levels. butes offered by advertisements in a brochure.
Choose an experimental design for gathering The eldwork for the qualitative research was
data. The experimental design is based on carried out in Segovia, Spain from November
statistical theory to combine levels of the attri- 2006 to February 2007. The use of university stu-
butes into alternative proles presented to dent samples for experimental research attracted
respondents. Whereas full factorial designs the attention of many researchers recently
often generate an impractically large number (Gallarza and Gil, 2006; Rewtrakunphaiboon
of combinations for evaluation, fractional and Oppewal, 2008). We used a focus group
factorial designs and block designs reduce with three smaller groups of eight students each
the number of combinations but with a con- based on a chosen destination of Punta Cana,
comitant loss of power. Santo Domingo. The sessions were conducted
Proles identied by the experimental design with two travel agency brochures, asking open-
are grouped into choice sets and presented to ended questions and recording answers. After
respondents. Proles can be presented to the sessions, we coded and classied answers
respondents individually, in respondent pairs to reveal relevant attributes. The data were
or in groups. coded using inductive category coding, consist-
Measurement of preferences. Individual pre- ing of labelling factors repeatedly found in text
ferences are discovered in choice-modelling (Spiggle 1994). Similar processes are used in
surveys by asking respondents to rank content analysis in the service literature (Tax
options, score them or choose those they et al., 1998), and the method is used extensively
prefer most. These methods of measuring in studies of consumer behaviour to identify
preferences correspond with variations of relationships within text (Spiggle, 1994).
modelling approaches. The attributes gathered in the focus groups
Estimation procedure. The model is adjusted were as follows: the image and aesthetics of
using Ordinary Least Square (OLS) regression the illustrations in the advertisement were an
or maximum likelihood estimation. important factor; aerial or panoramic images
were preferred, although well-lit, nocturnal
In this study, preference differences between photos were also viewed favourably; the size
hotel listings indicated differences in value of the illustrations and the hotels starred rating
Copyright 2012 John Wiley & Sons, Ltd. Int. J. Tourism Res. (2012)
DOI: 10.1002/jtr
R. Huertas-Garcia, M. Laguna Garca and C. Consolacin
were determining factors; ve-star establish- reported that the hotel listings position was
ments displayed in large photographs were irrelevant, we incorporated it anyway given
preferred; the structure of the buildings was the inuence product placement has on prefer-
very important; low, open buildings were ences (Cox, 1970; Chevalier, 1975). To deter-
preferred to high, square buildings, although mine high and low prices, we calculated the
there was no consensus concerning the size of average ranks of high and low prices from a
the hotel; accessibility and proximity to the sample of Santo Domingo hotels published in
beach and restaurants were important; the three brochures.
starred rating was indicative of quality, but no
importance was given to the ICTEs Q (Spanish
tourist quality symbol); subjects preferred the Quantitative research. For quantitative research,
price to be on the same page as the advertise- we propose a choice-based conjoint experi-
ment that described the hotel, although price ment of a sample of undergraduate students
range was not one of the most important who travel in groups. The underlying choice
attributes. A wide range of hotel services was surveys were the statistical design of the
preferred but not essential. No remarks were experiment, used to allocate all combinations of
made or attention paid to room amenities (e.g. factors and their levels to form a set of alterna-
air conditioning) or a hotels surroundings (e.g. tives to be used on the survey. An experimental
topography). design is the systematic arrangement of proles
Not all attributes must or can be considered; in matrices of coded values researchers use to
they must be chosen realistically and in a way describe attribute levels representing hypothet-
that is appropriate to the situation. Gustafsson ical alternatives of marketing options in the
et al. (1999) suggested a list of rules for choosing choice set. It is not an easy task; Kuhfeld et al.
attributes when evaluating service quality: (1994) pointed out that the best designs are
choose attributes that are important when the discovered when the researcher uses both
interviewees are buying, ones that can be modi- human design skills and computerized searches.
ed and ones used to compare with competitors. To design the experiment, we considered eight
Because there is no consensus in the literature factors: three quantitative variables (starred
about the elements encompassing each of the ratings, advertisement size and prices) and ve
positive and negative dimensions of value qualitative variables (names, descriptions of
(Woodruff, 1997), eight relevant attributes were hotel size, brochure illustrations, positions in
chosen for evaluation: the pamphlet and positions on the page). All
variables were coded as vectors (1, 1); for
Name (one short name Majestic Punta Cana factor 1, name, a positive sign indicated the long
and one long name Majestic Colonial Punta name Majestic Colonial Punta Cana Beach
Cana Beach Resort Golf Casino & Spa); Resort Golf Casino & Spa, and a negative sign
Hotels starred rating (four or ve stars); indicated the short name Majestic Punta Cana
Phrase used to describe the hotel size (24 and so on. Table 1 shows the factors and their
buildings with three oors, suites and junior codes. A full factorial design would have
suites; 659 suites and junior suites); required a 28 factorial experiment (i.e. 256
Picture in the brochure (a general view of the proles in the choice set). To avoid saturation
tourist building or a view of the swimming of information to respondents, we used a IV
pool); resolution design, 284
IV , a fraction one-sixth of
Size of the hotel listing (quarter or half page); the 28 full design. Resolution levels determine
Position in the brochure (right or left page); the degree of confusion generated by the experi-
Position on the page (top or bottom); ment among factors and interactions between
Prices (a low price of 277 or a high price them. A IV resolution design avoids confusion
of 499) between main effects and second-order interac-
tions, but two-factor interactions are mutually
The rst ve variables were gathered from confused. A complete overview of the design
the qualitative research, and the last three were of the experiment can be found in Box et al.
added by the researchers. Although subjects (2005) and Myers and Montgomery (2002).
Copyright 2012 John Wiley & Sons, Ltd. Int. J. Tourism Res. (2012)
DOI: 10.1002/jtr
Understanding Tourist Decisions
Copyright 2012 John Wiley & Sons, Ltd. Int. J. Tourism Res. (2012)
DOI: 10.1002/jtr
R. Huertas-Garcia, M. Laguna Garca and C. Consolacin
Variables Name stars Hotel size advertisement size left location on page Price 3 days Block variables
as independent. Each of the choices is called a odds model (or cumulative logit model) by
pseudo-observation because each complete using the PLUM process from SPSS.
ranking is only one observation formed by
multiple pseudo-observations depending on RESULTS AND DISCUSSION
the number of choice sets (Train, 2009). The
model considers that the rst option is chosen The results of estimation are shown in Table 4.
among seven proles; afterwards, the decision The ordinal logistic regression model has a de-
maker chooses the second option among the gree of adjustment usual in this type of study
remains six proles and so on. The probability (Cox and Snells pseudo R2 = 0.220; 2 log-likeli-
of any ranking of an alternative made by indi- hood = 160.768, p < 0.001) (Louviere et al., 2000).
vidual i can be expressed as a sequence of To determine whether the model offers adequate
probability multiplications: predictions, we used two log-likelihood values,
comparing two log-likelihood values for the
2 3 intercept-only model and the nal model with
J 6 7 the predictors. The chi-square statistic indicates
Y 6 expVij 7
PiUi1 > Ui2 > > Uij 6 7 whether the model offers signicant improve-
6PJ 7
j1 4 5 ment over the baseline intercept-only model.
expVik
kj The dependent variable had only ve of the
seven possible values, meaning the two exposed
(2) hotel listings were chosen between the rst and
Pq fth orders. Hence, the ordinal model estimates
where Vij b0 i1 bi xi e , Pi (Uij) only four cut-off values that separate the ve
measures the probability of this chosen order, response categories. Parameter estimates are
bi are the values of the slope of the vector for shown in Table 4, where four of the 11 effects
each main factor and e is an error term. An are signicant. These are B (starred ratings),
ordinal logistic regression model was used to H (size of hotel listing), E (prices) and B2 (the
adjust the full prole data in the proportional blocking variable). Whereas a variable in
Copyright 2012 John Wiley & Sons, Ltd. Int. J. Tourism Res. (2012)
DOI: 10.1002/jtr
Understanding Tourist Decisions
statistical design experiment can be inter- effects from the derivative model. Values are
preted singly if there is no evidence that it is similar to the previous analysis.
confused with another, B2 (the blocking vari- H (prices) had a negative impact on evalu-
able) is confused with several second-order ation by the sample. This result is in accord
interactions, B2 = AC = VG = DF = HE; inter- with the interpretation of price as sacrice; the
pretation is, therefore, imprecise. We suspect lower the price, the greater the customers
that the value of B2 responds signicantly to evaluation. Other more highly valued effects
the interaction between H (size of hotel listing) were as follows: B (starred ratings), E (hotel
and E (prices). listing size) and HE (interaction between price
To verify these results, we considered the and size of hotel listing). B (starred ratings)
analysis of only three factors, those which and E (advertisement size) had a positive inu-
demonstrated signicance after the rst experi- ence on the subjects evaluation. Respondents
ment of the three main effects (B, H and E). used both factors as external value cues; the
Because only three factors were signicant, a higher the starred ratings and the bigger the size
full factorial of 23 required eight proles drawn of the advertisement, the greater its evaluation.
from the previous design using 16 proles. This Moreover, the joint presence of HE (interaction
is possible because any fractional factorial between price and size of hotel listing) increased
design of resolution R may contain complete positive valuation. The dual role of price as
factorials in any sub-group of R 1 variables. sacrice and extrinsic cue of service quality is
Because the design is resolution R IV, it was appreciated when price interacts with a large
possible to congure a complete factorial advertisement. Thus, a negative evaluation repre-
design with these three variables. This design sented by high price accommodation is offset by
allows estimation of main effects and interac- a large advertisement. Likewise, the positive
tions without any confusion patterns and, conse- inuence generated by a large advertisement
quently, allowed direct interpretation. Ordinal increases with price.
logistic regression was used to evaluate
signicance of the effects. Table 5 shows the
model-tting information and the estimated CONCLUSIONS
value perception in the process of choosing a before purchasing is extremely useful for busi-
hotel from a travel agent brochure. Subjects nesses and managers. A travel agent brochure
used a reduced number of attributes to infer may be giving more prominence to elements that
perceived quality and the sacrice necessary do not determine perceived value of a tourist,
to acquire it. In this case, only three of the eight making communications ineffective and not
attributes were signicant: advertisement size contributing to a consumers decision. However,
(quarter or half page) and the hotels starred the validity of the attributes of perceived value
rating (4 or 5) as quality attributes, and price disappears with purchase. Fisher et al. (1994)
indicating sacrice. For this market segment argued that when consumers evaluate perceived
of students in their nal year of university, bro- service quality after purchase, they rarely
chure illustration was not signicant in spite of mention the criteria used for evaluation before
the results from qualitative research, neither purchase; when they do, such criteria are relegated
were position on page, nor name of to a lower hierarchic level to the one occupied
establishment, nor left versus right location, before purchase.
although results are in line with the qualitative This study has some limitations. Participants
research. Hotel size occupies an intermediate were from one university, disallowing statis-
position of importance; an increase in sample tical inferences from the sample to a popula-
size is necessary to verify the variables inu- tion. Some variables such as room quality and
ence on selection. hotel location were ignored as potentially rele-
Attributable to the second analysis using a vant attributes. The laboratory environment
full factorial design, it was possible to detect that afforded internal validity because of envir-
the second inductive factor of perceived quality onment control compromised external validity
consisting of the joint presence of two vari- because of an articial venue. The sample size
ables, advertisement size (quarter or half page) was also a limiting factor because it did not
and price. The two joint variables demonstrate allow us to determine the signicance of some
the dual role of price (Dodds et al., 1991), as a primary variables (e.g. hotel size), although the
sacrice people must make to buy goods or as model using ordinal logistic regression was
an extrinsic cue of service quality, when it inter- proposed because of its suitability to t small
acts with a large advertisement. Interaction samples. Designing this experiment, we
between price and hotel listing size showed a transformed a full fractional model into a
positive result when negative was expected, fractional one; the resulting design was orga-
suggesting some kind of compensatory effect. nized into blocks of two. This design only
These ndings have important implications for allows the estimation of main factors without
managers, specically that a higher price interactions. The design could be improved
expected to generate negative utility is offset by by taking advantage of multiple runs and
a larger advertisement to produce a positive using a fractional factorial with higher reso-
effect on a tourists nal decision. The increased lution such as R V, although there is the
cost of a larger advertisement could be recouped disadvantage of the cost of designing 32 bro-
by charging a higher price for the trip and chures instead of eight. The model helped us
hotel room. get closer to customers, identifying opinions
This study considers perceived value attributes and preferences; it was a simple procedure that
in a pre-purchase stage, a stage traditionally facilitated sequential research that enabled us
understudied. Knowing which aspects or attri- to continuously improve knowledge of the
butes determine perceived value of the product subject.
Copyright 2012 John Wiley & Sons, Ltd. Int. J. Tourism Res. (2012)
DOI: 10.1002/jtr
Understanding Tourist Decisions
Appendix A
Example of experimental travel agent brochure
Copyright 2012 John Wiley & Sons, Ltd. Int. J. Tourism Res. (2012)
DOI: 10.1002/jtr
R. Huertas-Garcia, M. Laguna Garca and C. Consolacin
Kotler P. 2000. Marketing Management. The Millen- Roehl W, Fesenmaier D. 1992. Risk perceptions and
nium Edition. Prentice Hall Inc: New Jersey. pleasure travel: an exploratory analysis. Journal of
Kuhfeld WF, Tobias RD, Garratt M. 1994. Efcient Travel Research 30(4): 1726.
experimental design with marketing applications. Scarpa R, Rose JM. 2008. Design efciency for non-
Journal of Marketing Research 31(4): 545557 market valuation with choice modeling: how to
Lee C-K, Yoon Y-S, Lee S-K. 2007. Investigating the measure it, what to report and why. The Australian
relationships among perceived value, satisfaction Journal of Agricultural and Resource Economics 52(3):
and recommendations: the case of Korean DMZ. 253282.
Tourism Management 28(1): 204214. Spiggle S. 1994. Analysis and interpretation of qualita-
Louviere JJ, Hensher, DA, Swait JD. 2000. Stated Choice tive data in consumer research. Journal of Consumer
Methods. Cambridge University Press: Cambridge. Research 21(3): 491503.
Mead R. 1988. The design of experiments. Cambridge:
Street D, Burgess L. 2007. The Construction of Optimal
Cambridge University Press; cited by Gilmour SG,
Stated Choice Experiments: Theory and Methods.
Trinca L A. 2006. Response surface experiments on
Wiley: Hoboken, NJ.
processes with high variation. In Response Surface
Tax S, Brown SW, Chandrashekaran M. 1998.
Methodology and Related Topics, Khuri AI (ed.). World
Customer evaluation of service complaint experi-
Scientic Publishing Co: New York.
ences: implications for relationship marketing.
Molina A, Esteban A. 2006. Tourism brochures:
Journal of Marketing 62(April): 6076.
usefulness and image. Annals of Tourism Research
33(4): 10361056. Teas RK, Agarwal, S. 2000. The effects of extrinsic
Moliner MA, Rodrguez RM, Callarisa Ll J, Snchez product cues on consumers perceptions of
J. 2005. Dimensionalidad del valor percibido quality, sacrice and value. Journal of the Academy
global de una compra. Revista Espaola de Investiga- of Marketing Science 28(2): 278290.
cin de Marketing 9(2): 135157. Teichert T, Shehu, E. 2007. Evolutionary conjoint. In
Munger JL, Grewal D. 2001. Effects of alternative Conjoint Measurement. Methods and Applications,
price promotional methods on consumers evalu- 4thedn , Gustafsson A, Herrmann A, Huber F
ation and purchase intentions. Journal of Product (eds). Springer: Berlin; 295320.
& Bran Management 10(3): 185197. Train K. 2009. Discrete Choice Methods with
Myers RH, Montgomery DC. 2002. Response Surface Simulation. 2ndedn , Cambridge University Press:
Methodology. John Wiley & Sons, Inc: New Jersey. Cambridge.
Naylor G, Frank, KE. 2001. The effect of price bundling Vriens M, Loosschilder, G, Rosbergen E, Wittink D.
on consumer perceptions of value. Journal of Services 1998. Verbal versus realistic pictorial representa-
Marketing 15(4): 270 281. tions in conjoint analysis with design attributes.
Nicolau JL, Mas FJ. 2006. Eleccin de la duracin Journal of Product Innovation Management 15(5):
vacacional: una aproximacin con modelos de 455467.
recuento. Revista Europea de Direccin y Economa Wicks BE, Schuett MA. 1991. Using travel brochures
de la Empresa 15(2): 99116. to target frequent travelers and big-spenders.
Ortega E, Rodriguez B. 2007. Information at tourism Journal of Travel & Tourism Marketing 2(3): 7790.
destinations. Importance and cross-cultural differ- Wittink DR, Cattin P. 1989. Commercial use of
ences between international and domestic tourist. conjoint analysis: an update. Journal of Marketing
Journal of Business Research 60(2): 146152. 53(3): 9196.
Ostrom A, Iacobucci D. 1995. Consumer trade-offs
and the evaluation of services. Journal of Marketing Woodruff BR. 1997. Customer value: the next source
59(1): 1728. for competitive advantage. Journal of the Academy
Rao, VR, Sattler, H. 2007. Measurement of price of Marketing Science 25(2): 139153.
effects with conjoint analysis: separating informa- Yamamoto D, Gill AM. 1999. Emerging trends in
tional and allocative effects of price. In Conjoint Japanese package tourism. Journal of Travel Research
Measurement. Methods and Applications, 4thedn , 38(2): 134143.
Gustafsson A, Herrmann A, Huber F (eds). Zeithaml VA. 1988. Consumer perceptions of price,
Springer: Berlin; 3146. quality and value: a means-end model and synthe-
Rewtrakunphaiboon W, Oppewal H. 2008. Effects of sis of evidence. Journal of Marketing 52(3): 222.
package holiday information presentation on Zhou Z. 1997. Destination marketing: measuring the
destination choice. Journal of Travel Research 47(2): effectiveness of brochures. Journal of Travel & Tourism
127136. Marketing 6(3): 143158.
Copyright 2012 John Wiley & Sons, Ltd. Int. J. Tourism Res. (2012)
DOI: 10.1002/jtr