Approximating Confidence Intervals for Factor Loadings

Abstract

Practical theoretic means for assessing the sampling variability of loadings estimated by exploratory factor analytic procedures have not been readily available in the absence of restrictive distributional assumptions. It has been necessary for researchers to interpret these point estimates (loadings) through the use of arbitrary rules-of-thumb. Under these conditions, loading interpretations may be problematic. A method is presented for exploiting information in the empirical data, collected for a study's primary goals, to approximate confidence intervals for factor loadings, The method appears generalizable across factor methods, numbers of extracted factors, and rotation criteria.

A strategic analysis for successful open source software utilization based on a structural equation model

“…The method is an attractive alternative to analytically derived results, because it does not need intricate derivations, and it usually relies on much weaker assumptions. The bootstrap method has been successfully used in a wide range of applications, among which are singular value decomposition-based methods (Chatterjee, 1984;Lambert et al, 1990Lambert et al, , 1991Efron and Tibshirani, 1993;Milan and Whittaker, 1995, p. 43;Timmerman and Ter Braak, in press In order to more thoroughly gauge the effects of the variables included in the model, the correlations between predictors and criterion variables were decomposed into the sum of their direct, indirect, and total effects (Tarí et al, 2007).…”

Applying the bootstrap to the multivariate case: Bootstrap component/factor analysis

“…CIs of the Procrustes-rotated pattern/structure coefficients and eigenvalues could be computed and could serve as an interpretation aid that takes sampling variability into account (Lambert, Wildt, & Durand, 1991). In 1999, the APA Task Force on Statistical Inference (Wilkinson & TFSI, 1999) strongly recommended the reporting of CIs.…”

“…When computing CIs with the bootstrap results, adjustments need to be made for biases inherent in the sample (Lambert et al, 1991).…”

A Factor Analytic Approach to Inferring Congestion Sharing Based on Flow Level Measurements

“…For Dataset2004 (significant loadings boxed) and Next, 95% confidence intervals for absolute values of eight rotated factor loadings are computed. When computing confidence intervals for factor loadings, one needs to take into account sign reversals of loadings and changes in the order of factors across bootstrap samples [38]. As such, we compute the confidence intervals of the absolute values of loadings.…”