Bifactor analysis - PennBBL/pncT1Bifactors GitHub Wiki

Author: Antonia Kaczkurkin
Date: 11/29/2018
Replicator: Tyler Moore
Goal: To generate bifactor scores from psychopathology items.

Data Preparation

1. Use the script `` to:
   a. Load the subject data
   b. Extract only the psychopathology items of interest and the bblids
   c. Save as a .csv file

Exploratory Analysis Overview

Exploratory factor analysis (EFA) is a statistical tool used to “discover” the structure underlying a set of item responses. It involves two steps:
A) initial extraction of m orthogonal, “unrotated” factors from a covariance matrix (m is determined by the researcher)
B) rotation of these factors to an interpretable structure

The number of factors to extract

The first step of factor extraction is to determine the number of factors necessary to account for a satisfactory amount of the variance in common among the variables (satisfactory being subjective).

Computational methods for the extraction of factors

Several methods exist for the extraction of factors from a correlation matrix (e.g. Principal-Axis, Maximum Likelihood, Least Squares).

Rotation

The ultimate goal of factor rotation is to identify interpretable and meaningful dimensions that account for, and explain, the relationships among test items.

Steps

1. Use the script `` to run an exploratory factor analysis on the data. This script will:
   a. Use the method for extracting factors.
   b. Rotate the factor solution using .
   c. The number of factors extracted was determined by .

Bifactor Analysis Overview