- 最后登录
- 2014-11-20
- 注册时间
- 2013-12-1
- 威望
- 0
- 金钱
- 103
- 贡献
- 40
- 阅读权限
- 20
- 积分
- 143
- 日志
- 0
- 记录
- 0
- 帖子
- 18
- 主题
- 6
- 精华
- 0
- 好友
- 0
![Rank: 3](static/image/common/star_level1.gif) ![Rank: 3](static/image/common/star_level1.gif) ![Rank: 3](static/image/common/star_level1.gif)
- 注册时间
- 2013-12-1
- 最后登录
- 2014-11-20
- 积分
- 143
- 精华
- 0
- 主题
- 6
- 帖子
- 18
|
罗老师,您好。在看文章中的验证性因子部分有几个疑问。假设文章有4个变量(1)有些文章的验证性因子分析结果对比了4因素模型和1因素模型的适配性指标,确定了4因素模型好。但有些文章展现了因子载荷,AVE及AVE的开方等说明因素间的建构效度。那么到底应该用什么方法呢?(2)在做这部分分析时,提到了Chi-Square different test,这个在spss或amos里具体怎么操作呢?
$ D- V# [ y, y& k) Z文章具体的验证性分析过程如下:
& t9 i8 Q0 ]0 ~0 w; h& K$ C% j* K9 |3 YTo test whether these four scales represent distinguishable constructs, I conducted a confirmatory factor ***ysis (CFA) on the individual responses. A four-factor model, normed fit index (NFI) = .93, comparative fit index (CFI) = .97, root-mean-square error of approximation (RMSEA) = .05, provided a better fit than a one-factor model, NFI = .60, CFI = .62, RMSEA = .19. I also performed a chi-square difference test to assess the significance of increased fit comparing the one-factor model to the four-factor model, △χ2= 935.57,p < .001. Thus, these results recommend the use of the four-factor model。
0 e: }+ u/ D8 a9 ]. P0 p2 M- j |
|