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

 
Scale Confirmation 
Confirmatory Factor Analysis (CFA) was conducted to verify seven-factor structure of the proposed 
experience
model, which was obtained from EFA as well as to explore whether significant modifications are 
necessary from EFA. Data from the second-split sample (N =199) was used for CFA. The proposed measurement 
model was specified by the relations between the observed variables and the unobserved latent variables (i.e., 
factors) through loadings of the observed variables and their error terms. All factors and observed variables were 
specified based on EFA results. The proposed measurement model consisted of seven factors and 39 observed 
variables. The proposed measurement model of the 
experience
was estimated using the AMOS 5.0 program. In past 
decades, a number of fit indices have been developed to evaluate the model fit. In this study, chi-square 
statistic/degree of freedom (< 3) as well as model fit indices such as comparative fit index (CFA > .9), non-normed 
fit index (NNFI > .9), and root-mean-square residual (RMSEA < .10) were examined to evaluate adequate model fit 
(Hu and Bentler, 1999; Kline 1998). Furthermore, the 
χ
2 
difference was used when comparing fit of an adequate 
model and other competing models. Models were evaluated by significance testing based on the difference in 
χ
2
and 
the degree of freedom between the two. 
The measurement model, which included seven factors and 39 observed variables, was tested and evaluated 
through the model specification procedure suggested by other researchers (Bollen, 1989; Joreskog and 
Sorbom,1993). Table 1 presents the overall fit indices for three models through specification and modification 
procedures.
 

---