Several important findings from the correlation tables are worth mentioning. Specificity
and intensity of attractions are highly and significantly correlated (r > 0.9; p < 0.0),
that is, in the areas where there are more first-class or specific attractions there are also a
higher number of total attractions. It seems, then, as if the presence of first-class attractions
fosters the development of other ‘complementary’ second-class attractions. This is
so in the case of both tourism zones and current destinations.
Moreover, both attributes, specificity and intensity of attractions, are also highly and
significantly correlated with numbers of beds in the case of tourism zones (r > 0.9;
p 0.8; p < 0.0) and to a minor extent to the number of
beds (r > 0.6; p < 0.0). This is a remarkable result as it proves that, at least in our case,
the more the tourism zones differ from the administrative regions, the higher is their
attractiveness.
Finally, as for the variety of attractions, we find no significant correlation with the rest
of the variables except for, to a certain extent, population (r > 0.6; p < 0.0). This points
at the idea that while all the zones have attractions belonging to most of the attractions’
categories, only the more populated zones offer the whole variety of attractions. Local
population living near tourism zones may consume some tourism products and attractions,
and therefore it is reasonable to assume that tourism companies may consider closeness
to highly inhabited areas as one of the factors when deciding their location. This is
particularly the case for some products and services, which, for example, need an important
investment in infrastructure when established, such as leisure parks or wellness
centres.
Types of tourism zones
The values of the attributes for each of the nine tourism zones are shown in Table 4.
Now the results of the hierarchical cluster analysis conducted with these variables
classify the nine zones into four types. Type I mountain tourism zone includes only one of