Travel Reference
In-Depth Information
with evaluations among a large number of attributes with different qualities that are diffi cult to
compare directly, which is often the case in destination choice.
An introduction to and empirical research on the application of AHP can be found in some
tourism studies (e.g. Deng
et al
. 2002; Crouch and Ritchie 2005; Calantone and di Benedetto
1991). Additionally this method was used by Hsu
et al
. (2009) as an analysis method to investigate
tourists' preferences of destination choice. A four-level AHP model with 22 sub-criteria on the
fourth level was used in this study. Compared to other tourism decision-making studies using
regression methods, it was able to provide the relative weights of a large number (22) of attributes
at one time. Furthermore, by clustering attributes into different levels, tourists only need
to evaluate the attributes with a similar nature, which makes the comparison easier. The
22 attributes estimated by Hsu
et al
. (2009) were initially divided into internal factors and
external factors, where the internal factors were further sub-divided into four categories
and external factors were divided into two categories. At each stage, respondents only need to
compare two attributes at the same level and within the same superior criterion.
Although the paired comparison for respondents at each stage is quite simple, there would
be a huge amount of workload if there were a large number of attributes within one category.
If for example there are nine attributes within the same superior criterion, then the respondents
need to complete 45 comparisons to make sure all the attributes are compared to each other.
Additionally, where there is a large number of alternatives the number of comparisons among
alternatives regarding each attribute's quality score would be too complex for respondents.
Furthermore, in the traditional AHP method, the pair-wise comparison is made using a nine-
point scale (1-9), which converts human preferences between available alternatives as equally,
moderately, strongly, very strongly or extremely preferred. In some real situations, respondents
might be reluctant or unable to provide exact numerical values to the comparison judgments.
Therefore, modifi cation and improvement of the traditional AHP approach concerning these
disadvantages are required.
In Hsu
et al
. (2009), the authors combined a fuzzy theory method with the traditional AHP
to reduce the workload of respondents, which allowed respondents to provide fuzzy judgements
instead of assigning precise comparison values. It is thus clear that a smart combination of
methods can be a good way to overcome the disadvantages of a single method and to make
estimations more effective.
How are attributes manipulated (choice heuristics) by tourists
to evaluate alternatives?
All the methods mentioned above are helpful for use in studies to gain more understanding
about which destination attributes are important to tourists and how much they are preferred.
As a matter of fact, in order to predict the fi nal choice of tourists' decision-making, we not only
need to know what attributes or factors are involved, but it is also necessary to understand the
choice heuristics that are applied by decision makers. The choice heuristic, or the evaluation
rules, refers to the way tourists use criteria to evaluate alternative destinations. For example,
some tourists (type A) may weigh every attribute carefully and select one destination with the
highest score whilst others (type B) may look at the most important attribute fi rst and keep
the alternative with the best performance. If there is a tie, then they would look at the second
important attribute and select the one with the best performance, until there is only one
destination left. As a matter of fact, even if the two types of tourists evaluate the same attributes
during their decision-making process their choices might be different because of the different
choice heuristics they applied.
