select_scalingind {MIIVefa}

Exploratory Factor Analysis Using Model Implied Instrumental Variables.

Description

This function selects a scaling indicator for an initial one factor model given a raw data frame matrix.

Usage

select_scalingind(data,
                 sigLevel = .05,
                 scalingCrit = "sargan+factorloading_R2",
                 correlatedErrors = NULL)

Arguments

data

A data frame, list or environment or an object coercible by as.data.frame to data frame. The most common application is to supply a data.frame.

sigLevel

The significance level threshold, default is .05.

scalingCrit

The criterion used to select the scaling indicators, default is 'sargan+factorloading_R2.'

correlatedErrors

The pairs of variables whose errors should be correlated in the model search procedure, default is NULL.

Details

All possible criteria for scalingCrit

order: uses the first appearing variable as the scaling indicator

sargan: uses the variable with the least number of significant sargans as the scaling indicator. if multiple variables have the same least number of significant sargans, chooses the first appearing variable.

R2: uses the variable with the highest R2 as the scaling indicator. if multiple variables have the same R2 value, chooses the first appearing variable.

factor loading: uses the variable with the most number of significant factor loadings as the scaling indicator. if multiple variables have the same most number of significant factor loadings, chooses the first appearing variable.

sargan_R2: uses the variable with the least number of significant sargans as the scaling indicator. if multiple variables have the same least number of significant sargans, chooses the one with higher R2. if still multiple options, chooses the first appearing variable.

sargan_factorloading: uses the variable with the least number of significant sargans as the scaling indicator. if multiple variables have the same least number of significant sargans, chooses the one with more significant factor loadings. if still multiple options, chooses the first appearing variable.

sargan_factorloading_R2: uses the variable with the least number of significant sargans as the scaling indicator. if multiple variables have the same least number of significant sargans, chooses the one with more significant factor loadings. if still multiple options, chooses the one with higher R2.

factorloading_R2: uses the variable with the most number of significant factor loadings as the scaling indicator. if multiple variables have the same nist number of significant factor loadings, chooses the one with higher R2. if still multiple options, chooses the first appearing variable.

factorloading_sargan: uses the variable with the most number of significant factor loadings as the scaling indicator. if multiple variables have the same nist number of significant factor loadings, chooses the one with less significant sargans. if still multiple options, chooses the first appearing variable.

factorloading_sargan_R2: uses the variable with the most number of significant factor loadings as the scaling indicator. if multiple variables have the same nist number of significant factor loadings, chooses the one with less significant sargans. if still multiple options, chooses the one with higher R2.

sargan+factorloading: uses the variable with the least sum of significant sargans and non-signficant factor loadings. if multiple variables have the same least sum of significnat sargans and non-significant factor loadings, chooses the first appearing variable.

sargan+factorloading_R2: uses the variable with the least sum of significant sargans and non-signficant factor loadings. if multiple variables have the same least sum of significnat sargans and non-significant factor loadings, chooses the one with higher R2.

Value

A vector indicating the variable (column name) that is suggested to be the scaling indicator for the initial one factor model.