calculateStrictSb: calculateStrictSb

In umx: Structural Equation Modeling and Twin Modeling in R

calculateStrictSb

Description

calculateStrictSb calculates the Satorra-Bentler (2010) scaled difference chi-square test (the "strictly positive" design relative to SB-2001, which can yield a negative scaling constant c_d). The returned statistic is only interpretable when both models share the same WLS moments, weight matrix, and N, and when the freer model's optimized discrepancy is not larger than the nested model's (F_{nested} \ge F_{base}). Non-monotone F (typically failed optimization) returns NA with a warning.

Usage

calculateStrictSb(baseModel, nestedModel)

Arguments

baseModel	The base model (more parameters)
nestedModel	The nested model (fewer parameters)

Value

A named vector containing strictSbChisq, deltaDf, scalingFactor, and pValue. May carry attribute nonMonotoneF = TRUE when \Delta F < 0.

umx documentation built on Oct. 11, 2026, 5:07 p.m.