The Frontal Assessment Battery 20 years later: normative data for a shortened version (FAB15)

Objective: The Frontal Assessment Battery (FAB) is a neuropsychological tool largely used to assess executive functions. Prior studies found a marked ceiling effect for the prehension behavior subtest (PBT) in healthy and clinical populations. Aims of the present study were (i) to examine the psychometric properties of the FAB without the contribution of PBT and (ii) to provide normative data for a revised version of the FAB after exclusion of PBT (FAB15). Methods: The normative sample included 1,187 healthy participants. PBT had near-zero variance, poor content validity, and no discrimination power. Internal consistency increased when PBT was excluded. We assessed the FAB15 factorial structure, interrater, and test-retest reliabilities. Normative data  for the FAB15 were extracted through a regression-based procedure according to sex, age, and education. Results: The principal component analysis revealed a single "executive factor" or alternatively a bifactorial solution reflecting the different degree of discriminative capability vs. difficulty of the subtests. The FAB15 demonstrated excellent interrater and test-retest reliabilities. Regression analysis showed that sex (lowly educated women < lowly educated men), higher age, and lower education affected FAB15 score. Accordingly, three grids for adjustment of raw scores (men, women, and both) were constructed. The cut-off was fixed at the non-parametric outer tolerance limit on the fifth centile (9.36, 95% CI). Conclusion: The observation of a ceiling effect in healthy subjects makes PBT not suitable for inclusion in a neuropsychological battery. The FAB15 may successfully replace the conventional FAB as a more severe and valid short screening tool to assess executive functioning.