factload {DandEFA}

R-mode Exploratory Factor Analysis

Description

This function produces a factor loading matrix given by the specified factor extraction method, number of factors and rotation method. R-mode Exploratory Factor Analysis is based on the factorization of the variables. See "References" for more details.

Usage

factload(data, cormeth = "spearman", data.cor, 
         method = "pc", nfac = 1, rotation = "varimax")

Arguments

data

Dataset to be analyzed for R-mode Exploratory Factor Analysis

cormeth

Correlation coefficient. Type polycor for polychoric correlation coefficient (see polychor), for else see cor. If data is NULL, and data.cor is specified; cormeth will be ignored.

data.cor

Given correlation matrix. However, if data is specified, data.cor will be replaced.

method

Factor extraction method. Type pc for Principal component factor analysis, mle for Maximum likelihood estimation and prax for Principal axis factor analysis (see "References")

nfac

The number of factors

rotation

Rotation method. Type varimax for orthogonal and promax for oblique rotation.

Value

Returns a loadings class factor loading matrix

References

Reyment, R., Joreskog, K.G., 1993. Applied Factor Analysis in the Natural Sciences. Cambridge University Press, Cambridge, United Kingdom.

Klinke, S., Wagner, C., 2008. Visualizing exploratory factor analysis models, in: Paulo, B. (Ed.), Compstat 2008: Proceedings in Computational Statistics: 18th Symposium Held in Porto, Portugal, 2008.

Harman, Harry H., and Wayne H. Jones. "Factor analysis by minimizing residuals (minres)." Psychometrika 31.3 (1966): 351-368.

Examples

data(timss2011)
timss2011 <- na.omit(timss2011)
factload(timss2011,nfac=5,method="prax",cormeth="spearman")
factload(timss2011,nfac=8,method="pc",cormeth="pearson",rotation="promax")
timsscor <- cor(timss2011)
factload(data.cor=timsscor,nfac=10,method="mle")