R 中的因子分析
我试图更好地理解 FA,希望你能看看这个,我最大的问题是如何在 R 中解释 FA 模型。
我的结果如下所示: 我应该关注结果中的哪些值以及 FA 分析的良好指标是什么?
Call:
factanal(x = m2, factors = 2)
Uniquenesses:
v1 v2 v3 v4 v5 v6 v7 v8 v9 v10 v11 v12
0.005 0.324 0.344 0.092 0.084 0.128 0.271 0.272 0.398 0.384 0.540 0.472
Loadings:
Factor1 Factor2
v1 0.847 0.527
v2 0.818
v3 0.733 0.344
v4 0.938 0.169
v5 0.949 0.125
v6 0.825 0.437
v7 0.701 0.488
v8 0.646 0.557
v9 0.467 0.619
v10 0.665 0.417
v11 0.525 0.429
v12 0.581 0.436
Factor1 Factor2
SS loadings 5.905 2.780
Proportion Var 0.492 0.232
Cumulative Var 0.492 0.724
Test of the hypothesis that 2 factors are sufficient.
The chi square statistic is 410.82 on 43 degrees of freedom.
The p-value is 1.59e-61
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(3)
我发布了 R 中的因子分析示例性格测试的因素结构。它展示了如何提取您可能需要的一些常见信息(例如,公共性;因素数量的检验;因素解释的方差;轮换等)。
I posted an example factor analysis in R looking at the factor structure of a personality test. It shows how to extract some of the common information that you might want (e.g., communalities; tests of number of factors; variance explained by factors; rotations; etc.).
一般来说,对于 FA,您无法直接解释因子载荷,因为它们不是唯一的(旋转问题)。除此之外,我讨厌听起来像心理学家(统计学家笑话......),但你的 p 值很低!
In general, with FA you cannot directly interpret the factor loadings because they are not unique (rotation problem). Other than that, I hate to sound like psychologist (statistician joke...), but you have a low p-value!
因为这里没有可重现的示例,而只是输出。我将为大家提出下一步全民教育的建议。
在这里我认为你需要验证你的模型的可靠性。一般来说,推荐使用
alpha()
和splitHalf()
函数,它们位于psych
包中。如果您发现模型的可靠性都大于 0.8,那么幸运的是您可能会得到一个好的模型。
DataCamp让您更深入。
Because here there is not a reproducible example but just an output. I will give the suggestion for the next step of EFA for you.
Here I think you need verify the reliability of your model. Generally,
alpha()
andsplitHalf()
functions are recommended, which are in thepsych
package.If you find your model's reliabilities are both larger than 0.8, fortunately you may get a good model.
There is an minimal examples on the DataCamp for you to go deeper.