Potential age bias in the NEO-PI-R: potential differential item functioning in older versus younger adults.

Geriatric researchers and clinicians often have to
deal with a lack of valid personality measures for older age
groups (e.g., Mroczek, Hurt, & Berman, 1999; Zweig 2008),
which hampers a reliable assessment of personality in later
life. An age-neutral measurement system is one of the basic
conditions for an accurate personality assessment across the
lifespan, both longitudinally and cross-sectionally. In the
present study, we empirically investigate the age-neutrality
of one of the most widely used personality measures (i.e.,
the NEO PI-R (Costa & McCrae, 1992)), by examining
potential Differential Item Functioning (DIF). Overall,
results indicate that the vast majority (92.9 % at domainlevel
and 95 % at facet-level) of the NEO PI-R items was
similarly endorsed by younger and older age groups with the
same position on the personality trait of interest, corroborating
the NEO PI-R's age neutrality. However, Differential
Test Functioning (DTF) analyses revealed large DTF for
Extraversion, and facet A6 (Tender-Mindedness). Results
are discussed in terms of their implications for using the
current format of the NEO PI-R in older aged samples.