Microsoft Word 2007 ichrie conference Proceedings Final-Final 06-06-07. doc

 
Canonical Correlation Analysis
Canonical correlation analysis was used to reveal the extent of the relationship and identify the existing 
variates between the two sets of constructs: quality of tourism experience and perceived destination competitiveness. 
Due to the fact that both sets of variables have a large number of items, the composite mean score of the items 
loading on the factors was calculated to be used as the construct score based on the above factor analysis results. The 
perceived destination competitiveness variables were considered the dependent variates, whereas the variables of 
quality of vacation experience were considered the independent variates.
The canonical correlation analysis results showed that three out of five canonical functions were 
statistically significant at the .001 probability level (Table 1). The overlapping variances between the pairs of 
canonical variates, which are equal to the squared canonical correlation for each function, ranged between 12.1% 
and 65.0%. This finding means that the first canonical variate explained 65.0% of the total variance shared by the 
two sets of variables (i.e., quality of vacation experience and destination competitiveness). The second function 
explained 22.7% of the leftover variance, and the third function explained 12.1% of the residual variance after being 
extracted by the first two functions. The redundancy index showed that 40.2% of the quality of vacation experience 
variance could be covered by perceived destination competitiveness variables, whereas 38.9% of the perceived 
destination competitiveness variance could be explained by quality of vacation experience items (Table 1).
Table 1.