Examining the Item-Level Factor Structure of the Behavior Rating Inventory of Executive Function-Adult Version Within a Traumatic Brain Injury Sample

BRIEF-A: Factor structure and measurement invariance across self-report and informant forms

OBJECTIVE

This study analyzes the psychometric properties of the two forms (self-report and informant) of the Behavior Rating Inventory of Executive Function-Adult version (BRIEF-A) in a sample of healthy Portuguese adults.

METHOD

The participants were 608 adults, 304 of whom answered the self-report form (ages 18-59; 137 male and 167 female) and 304 who answered the informant form (ages 18-70; 110 male and 194 female).

RESULTS

The internal consistency for the indexes and the Global Executive Composite was very good (≥.90), whereas for the scales they were either acceptable (between .70 and .80) or good (≥.80), except for the Inhibit scale in both forms. Confirmatory factor analysis was used to test seven competing factor models for each of the forms. The original two-factor model (Behavioral Regulation and Metacognition Indexes) showed a slightly better model fit than the three-factor model (Behavioral Regulation, Emotional Regulation, and Metacognition Indexes) in both forms. The multiple-group analysis of the two-correlated-factor model across forms was supported (configural, metric, and partial scalar invariance).

CONCLUSIONS

Overall, the BRIEF-A showed adequate psychometric properties, suggesting that it is a useful instrument to assess everyday executive functioning in healthy Portuguese adults.

Patient-Reported Outcome of Functional Cognition for Traumatic Brain Injury: Item-Level Psychometrics

A patient-reported functional cognition measure that can bridge neuropsychological tests and ecological validity is lacking. The purpose of this study was to present the initial item-level psychometrics of a functional cognition item bank, the Functional Cognition-Patient-Reported Outcome (FC-PRO). Confirmatory factor analysis, Rasch analysis, and convergent validity were conducted to establish item-level psychometrics of the FC-PRO item bank. Four of the six domains met essential unidimensionality criteria; all domains met rating scale criteria; domains had relatively few misfitting items (3%-13%); person reliability (.8-.92), and person strata (2.97-4.29) were satisfactory; the domains showed convergent validity for five of six domains. The FC-PRO showed adequate initial item-level psychometrics for researchers/clinicians to measure specific functional cognitive domains in patients with traumatic brain injury (TBI). This study provides the foundation for the development of future short forms, computer adaptive tests, and keyform recovery maps of patient-reported functional cognition to facilitate clinical utilization.