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

Overall Results of Canonical Correlation Analysis (N=353) 
Variate 
number 
1 2 3 4 5 
Canonical correlation 
.806 .476 .348 .158 .101 
Eigenvalue (Canonical R
2
) 
.650 .227 .121 .025 .010 
Wilk’s lambda significance 
.000 .000 .000 .015 .060 
% of variance traced 
Vacation experience 
.549 .096 .154 .105 .095 
Cumulative 
% 
.549 .645 .799 .904 .999 
Destination 
competitiveness 
.537 .077 .154 .139 .093 
Cumulative 
% 
.537 .614 .768 .907 1.00 
Redundancy % 
Vacation experience items 
.357 .022 .019 .003 .001 
Cumulative 
% 
.357 .379 .398 .401 .402 
Destination 
competitiveness 
items 
.349 .017 .019 .003 .001 
Cumulative 
% 
.349 .366 .385 .388 .389 
In order to determine which variables were most important in a given pair of canonical variates, 
standardized canonical coefficients (canonical weights) and canonical loadings (i.e., correlations between each 
variable and canonical variate) were checked. Canonical loadings are used to measure the linear correlation between 
a variable in the dependent or independent set and the set’s canonical variate (Hair et al., 1998). In this study, only 
variables with a canonical loading greater than .40 were considered for the interpretation of the variates (Table 2).

---