Principal components analysis of the Neurobehavioral Symptom Inventory in a nonclinical civilian sample

ABSTRACT

The study examined the component structure of the Neurobehavioral Symptom Inventory (NSI) under five different models. The evaluated models comprised the full NSI (NSI-22) and the NSI-20 (NSI minus two orphan items). A civilian nonclinical sample was used. The 575 volunteers were predominantly university students who screened negative for mild TBI. The study design was cross-sectional, with questionnaires administered online. The main measure was the Neurobehavioral Symptom Inventory. Subscale, total and embedded validity scores were derived (the Validity-10, the LOW6, and the NIM5). In both models, the principal components analysis yielded two intercorrelated components (psychological and somatic/sensory) with acceptable internal consistency (alphas > 0.80). In this civilian nonclinical sample, the NSI had two underlying components. These components represent psychological and somatic/sensory neurobehavioral symptoms.

KEYWORDS:

• Concussion
• minor head injury
• postconcussion
• traumatic brain injury

Acknowledgments

This research was approved by the Queensland University of Technology University Human Research Ethics Committee (ref: QUT HREC 1300000224). The authors thank the students who assisted with the data collection, Dr. Shannon L. Edmed who contributed to early discussions about this project, and Dr. Sherrie-Anne Kaye who assisted with the production of the revised manuscript.