<![CDATA[Interdependency analysis (canonical correlation analysis) intends to determine how much influence among the variables comprising two groups of variables (set variable) reciprocal between demand variables non-functional group member to request a non-functional group non-member. By using a sample of 45 respondents which were valid and reliable, normally distributed, homogeneous, numeric and no autocorrelation, the resulting data consist of 2 groups (2 sets), each group consisting of three indicator variables (manifest) variables, namely Mode Effect(M1 and NM1), Prestige effect (M2 and nm2) and the Veblen effect (M3 and NM3) with the hypothesis that the variable demand affects Non-functional Member Non-Member. Data were analyzed with the model of canonical correlation and MANOVA statistical test at alpha level of 5% (1 and 2-way), with the tools SPSS ver.21, showed that the effect of the variable group (based on eigenvalue) Member of the Non-Member is larger than the effect of non-Member to Member in 2 of 3 Root generated. These results indicate that the effect of Mode, prestige, and Veblen affect to non-member and statistically significant.]]>
Copyrights © 2013
