De Fruyt and De Clercq (this volume) and Sellbom (this volume) raise important issues surrounding the use of a five-factor model (FFM) of personality to conceptualize, assess, and diagnose personality disorder including how one includes a measure of impairment, the level of abstraction that is optimal (domains vs. facets), and the need for the development of formal test manuals, normative data, and means for identifying non-credible responding. In this response, the authors note their agreement with De Fruyt and De Clercq regarding the importance of assessing impairment but note that (a) it is already included to a large degree in the assessment of pathological FFM traits and (b) that they would prefer an approach that focuses explicitly on difficulties in occupational and social functioning. As to Sellbom’s comments suggesting that further work is necessary with regard to the development of test manuals, normative databases, and measures of valid responding – the authors agree and note some of the obstacles including the need for funding for the collection of normative data (and what consensus as to the kind – clinical only; community only; both) and development of test manuals.  As to measures of credible responding, they note that many of these exist already for the family of FFM PD scales that they helped create and are aware of similar efforts for other popular measures of pathological traits.

Invalid responding is an important consideration in mental health assessment. Given that most assessment data are gathered from self-report methods, accurate diagnostic and clinical impressions can be compromised by various forms of response bias. In this chapter, we review the ways in which evaluations of psychopathology, neurocognitive symptoms, and medical/somatic presentations can be compromised due to noncredible responding and invalidating test-taking approaches. We cover a variety of strategies and measures that have been developed to assess invalid responding. Further, we discuss evaluation contexts in which invalid responding is most likely to occur. We conclude with some remarks regarding cultural considerations as well as how technology can be incorporated into the assessment of response bias.