Academic journal article Psychologische Beiträge

On Faking Personality Inventories

Academic journal article Psychologische Beiträge

On Faking Personality Inventories

Article excerpt


The results of three experiments are reported, which prove the testees' profit-aimed faking behavior in answering personality inventories. The subjects were 60 students of psychology, 426 participants of a psycho-therapeutic introductory course, and again 151 students of psychology. The personality inventories used consisted of an unpublished 8-scales inventory for measuring service-orientation, the MBTI, and a forthcoming 4-scales inventory based on the well-known sedimentation hypothesis. The item response format was a four-category graded, a dichotomous, and an analogue scale response format, respectively. Every experiment yielded at least one scale which appeared to be faked (good), in spite of the fact that the threatened negative consequences were fairly minor. In the case of using analogue scale as response format, the subjects accidentally faked bad. It is concluded that psycho-diagnosis based on personality inventory might better not to be risked.

Key words: faking good, item response format, personality inventory, Rasch model

1. Introduction

This paper presents a detailed and in-depth investigation of testees' answering behavior in response to personality inventories. It is based on three experiments using different inventories and different response formats and partly applies to a large number of subjects.

Obviously, personality inventories assume that respondents answer the questions truthfully; however, there is never any guarantee that they actually do so. It is highly likely that many respondents are unwilling to tell the truth about themselves - even if they know the truth! As a consequence, this critical point of view gave rise to increased usage of projective tests, as these may not be faked as easily as responses to personality inventories.

Previous research findings demonstrated, of course, that individuals can distort their responses to personality inventories when directed to do so (e.g., Noll, 1951; Wesman, 1952), and in particular, when directed to answer in a socially desirable (SD) direction (cf. Edwards, 1970). Hence, already some decades ago a debate arose as to whether measures of personality merely assess the ability to detect responses that are socially acceptable or unacceptable. As a result, much effort has gone into attempts to control the social desirability response set, for instance by the development of separate social desirability scales or infrequency scales (cf. Block, 1965). However, there is almost no confidence in the effectiveness of these efforts.

Since the early research work cited above, many papers have dealt with the problem of faking when testees are confronted with personality inventories; for instance, the well-known study of Hacker, Schwenkmezger and Utz (1979) demonstrated that 5 of 12 and 5 of 16 scales within the frequently used German Freiburger Personlichkeitsinventar and within the 16 PF differ in their scores significantly, depending either on being administered in the standard manner or with SD-instruction, the latter inducing the testee to answer towards social desirability. An extensive overview is provided by Rogers (1997). Ones, Viswesvaran and Reiss (1996) and Viswesvaran and Ones (1999), respectively, even provide some kind of meta-analysis.

Yet it appears that almost nobody pays sufficient attention to the facts: Practitioners apply personality inventories daily and frequently base their personality assessments exclusively on them. Therefore, this paper attempts once more to put forward a convincing experience-based argument against personality inventories, now that their experimental design is extremely facetted.

2. Experiment I

In the first experiment 60 students of psychology were randomized and underwent two different test instructions. While the control group was tested by a personality inventory in the standard way - "please always select that response which describes your own situation best: strongly agree, agree, disagree, or strongly disagree" - the experimental group was tested after being asked to imagine the challenge of some job placement situation: ". …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.