The FFPI

For an extensive description and account of the Five Factor Personality Inventory, see Hendriks (1997); Hendriks, Hofstee & De Raad (1999a, 1999b, 2011).

The FFPI differs from other personality questionnaires on a number of points:

a. Coverage of the domain of personality traits: for his dissertation, Brokken (1978) collected and analyzed over 1200 trait adjectives. His point of departure was the ‘lexical’ hypothesis, which holds that the language community finds words for all traits that are of importance. A consideration in the background is that the psychology of personality has to rely on the human judge: objective ‘measurement’ of personality is an illusion.

b. Concreteness: in the framework of his dissertation research on pragmatic aspects of everyday personality language, De Raad (1985) noticed that trait adjectives as such hardly occur in common usage. Consequently, for the later construction of the FFPI (Hendriks, 1997) clusters of adjectives were ‘translated’ into a large collection of brief, concrete, behavior descriptive sentences.

c. Intersubjectivity: Hofstee (1994) offers an elaborate argumentation for multiple judges in personality diagnostics, and against (mere) self-report. A central argument is that judges agree insufficiently, so that others who know the person well are needed to reach reliability. The FFPI embodies that principle because the items are formulated in the third person singular (‘someone who...’). A second central argument for calling upon others as judges is that someone's personality is a factor in his or her relations with the close social environment.

d. Efficiency: whereas in traditional questionnaires, 10 to 20 items ar needed for an internally consistent score, the FFPI satisfies that requirement with hardly more than two items per trait. Efficiency is reached, first, through scoring based on principal components, so that each item contributes (more or less) to each trait scale; second, through application of the Abridged Big Five Circumplex (Hofstee, De Raad, & Goldberg, 1992) network model, which descibes traits as combinations of each other (instead of specifications of more general traits).

e. Bipolarity: traits come in pairs of opposites: talkative-silent, reliable-unreliable. Their scale is thus bipolar, with a natural zero point at which the trait passes into its opposite. Reporting on the traditional relative scale, with the mean score as a zero point, raises misunderstandings, as that mean tends to lie on the socially desirable side of the zero point: with a relative scale, someone with a positive score between the mean and the midpoint ends up with a negative score. The FFPI reports in terms of deviations from the zero point (see, Hofstee & Hendriks, 1998)

The FFPI-III differs from the earlier versions on minor details only; however, it is a public instrument. That popularization is to express that this kind of psychology is to serve primarily the person himself or herself. Use by order of others, for decisions about the person, leads to strategic responding (faking, malingering). Open access does not take away that for optimal use a professional context is needed, like in coaching, psychological consultation, and the like.

References: