Neonatal Behavioral Assessment Scale (NBAS): Confirmatory factor analysis of the six behavioral clusters

The Neonatal Behavioral Assessment Scale (NBAS) is a widely used in the neurobehavioral assessment of neonates in clinical practice and research. Lester's data reduction system for the NBAS items is the most often used in research, but the few factor analytic studies carried out with it leave gaps in its validation. The current study aimed to test and compare (a) the factorial structure of the Lester's data reduction system for the NBAS and (b) an alternative data reduction system, slightly modified from Lester's system. The NBAS was administered to 196 healthy Portuguese full-term infants (51% male) in the first 72 h of life (M = 43.63 h). Construct validity of the data reduction systems was tested through confirmatory factor analysis (CFA). Lester's original system was compared to three alternative models, two of which included a revision of the scoring rules for three items and the exclusion of five items. The CFA generally supported the six-factor structure. However, fit indices for Lester's original model were only fair. An alternative, revised model with a second-order factor - Self-Organizing System - demonstrated a better fit. The results provide evidence to support a modified form of Lester's six behavioral clusters as a data reduction model for the NBAS items.