简体繁体 English

R中fa（）和factanal（）函数之间的区别

[英]Differences between fa() and factanal() functions in R

原文 2018-05-28 21:50:24 7 1 r/ statistics/ components/ factors

I'm performing an exploratory factor analysis on my dataset, composed of 23 variables (which are questions asked to people). 我正在对数据集进行探索性因素分析，该数据集由23个变量（向人们提出的问题）组成。 I got 1777 observations for these variables and I would like to investigate the latent factors that can explain these variables. 我对这些变量获得了1777个观测值，我想研究可以解释这些变量的潜在因素。 Therefore, I used first the function fa() and then I tried with the function factanal() which doesn't give the same results for which I have difficulties to interpret. 因此，我首先使用了fa()函数，然后尝试使用factanal()函数，但给出的结果却难以解释。

Does anyone can explain me the differences between fa() and factanal() functions please? 有谁能向我解释fa()和factanal()函数之间的区别？

1 个解决方案

Actually, after analysing the help in R, I found the main differences which are: 实际上，在分析了R中的帮助之后，我发现主要区别在于：

factanal performs a maximum-likelihood factor analysis on a covariance matrix or a data matrix while fa() function is a more general function in the sense that it proposes different fit methods such as Ordinary least square regressions (OLR). factanal对协方差矩阵或数据矩阵执行最大似然因子分析，而fa（）函数是更通用的函数，因为它提出了不同的拟合方法，例如普通最小二乘回归（OLR）。 fa() also permits to draw a Factor Analysis diagram in order to retrieve visually which factors explain which variables with the different loadings showed. fa（）还允许绘制一个“因子分析”图，以便从视觉上检索哪些因子解释了哪些变量显示了不同的载荷。

In conclusion, I would say that fa() permits more flexibility than factanal() 总而言之，我想说fa（）允许比factanal（）更大的灵活性