One hundred and twenty five adult job applicants completed a computer-administered personality questionnaire (EPP) and two intelligence tests; one of crystalized (Wonderlic) and one of fluid intelligence. There were negative correlations between dissimulation on the personality questionnaire and performance on the two intelligence tests.
It is not entirely clear what faking, lying and dissimulation scales on personality tests actually measure (Furnham, 1986). For instance, the lie (L) score on the Eysenck Personality scales may supply not only a "detection" index but also another personality variable measuring social conformity or naivety. Paulus (1991) has distinguished between two types of socially desirable responses: selfdeception and impression management. The former measures an honest, but overly positive self-presentation, while the latter measures self-presentation tailored to a particular audience. Many researchers have claimed that the various dissimulation measures actually tap into psychological constructs which are of interest in their own right. There seems to be a high degree of consensus as regards the fact that socially desirable responding (i.e., dissimulation, lying) is trait-like in the sense that it is fairly stable over time (Furnham, 1986) though it rarely features in research as an independent trait variable.
Most research on dissimulation has been conducted within the personality area, where it is assumed that particular tests are vulnerable to faking good.
Much less has been done in the area of ability, though many have considered the relationship between personality and intelligence (Brebner & Stough, 1995; Eysenck, 1971; Matthews & Dom, 1995; Rawlings & Camie, 1989; Robinson, 1985; Saklofske & Kostura, 1990). An exception is the work of Iverson (Iverson, 2001; Iverson, Slick, & Franzen, 2000; Iverson, Franzen, & McCracken, 1992) and Slick (Slick, Hopp, Strauss, & Hunter, 1994) who have examined malingering in ability tests. These authors have both specified criteria for "diagnosing" malingering and proposed a number of methods to detect it.
Furnham, Forde, and Cotter (1998a) examined the relationship between the big three Eysenckian personality factors and test-taking style (dissimulation, time taken to complete the test, number of "can't decide" responses) on two intelligence tests; the one of crystalized, the other of fluid intelligence. There seemed to be evidence that both sets of factors were important such that stable introverts who responded quickly with low dissimulation scores did better on the tests.
This is a partial replication of the study mentioned above focusing on the role of dissimulation. People who score high on dissimulation tests are thought to be manifesting a mixture of social conformity and need for social approval. These dispositions are very likely to lead to the well-established faking profile, namely high extraversion, but low neuroticism and psychoticism scores. It also may be expected that there would be a negative correlation with ability tests, as dissimulators would be particularly eager to impress with their intelligence.
There were 125 participants in this study, of whom 80 were males and 45 were females. Their mean age was 32.73 years (SD = 7.72). They were job applicants for a middle-management post, who were required to complete a battery of inventories and tests.
The Eysenck Personality Profile (EPP) (Eysenck, Barrett, Wilson, & Jackson, 1992). The EPP is a 420-item questionnaire measuring 21 primary and three super-traits. The test is computer administered. Each item has a three-point response scale: yes, no and can't decide. In addition to these 24 scores, the test provides scores on dissimulation (or lie scale), response latency (total amount of time to complete the test recorded in minutes and seconds), and total can't decide responses. The EPP has long been subjected to psychometric assessment (Costa & McCrae, 1995; Eysenck et al., 1992; Wilson & Jackson, 1994) and has satisfactory evidence of both reliability and validity.
The Wonderlic Personnel Test (WPT) (Wonderlic, 1992). This is a 50-item, timed test (12 minutes) that measures crytalized intelligence. Scores range from 0 to 50 and the test has impressive norms. Items include word comparisons, disarranged sentences, number comparisons and series analysis of geometric figures and story problems requiring mathematical and logical solutions. The test manual reports on abundant evidence of construct and concurrent validity, as well as reliability. The WPT correlates with the WAIS-R at gamma = 0.92.
The Baddeley Reasoning Test (Baddeley, 1968). This is a 60-item, timed test (3 minutes) that measures fluid intelligence. Scores range from 0 to 60. The test is based on simple grammatical transformations that have to be answered "true/false"; for example, A precedes B-AB; A does not follow B-AB. It provides two scores: the number of questions attempted and the number of correct responses. The test has often been used in cognitive experiments to obtain a quick and reliable measure of verbal or cognitive ability (Farnham, Gunter, & Peterson 1994: Hammerton, 1969).
All participants completed the tests as part of an assessment center designed by a psychological consultancy for a particular company. Participants were administered all tests strictly in accordance with prescribed instructions. Participants were aware that their scores would be used in the final selection decision.
Examination of the means suggested the population was about one standard deviation above the norm for both ability tests. Further, the means and variances for the personality scores were in the normal range (Jackson, Furnham, Forde, & Cotter, 2000).
Three regressions were performed, whereby scores on the Wonderlic and Baddeley tests (correct responses and number attempted) were regressed on the three Eysenckian factors; dissimulation, number of can't decide responses, and time taken to complete the EPP. The results of the three regressions are shown in Table 1. For the Wonderlic measure of crystalized intelligence, the three Eysenckian factors were not significant but two of the test style variables were.
Those with higher lie scores and those who took longer to complete the personality test did less well on the Wonderlic intelligence test. This accounted for 16 percent of the variance. For the two measures taken from the Baddeley test it was only the lie scale that predicted test scores. Higher scores were associated with lower scores and fewer items attempted on the Baddeley test.
Results confirmed the basic hypothesis of this paper that dissimulation is correlated with intelligence as well as with personality test scores. High lie scores are associated with low intelligence test scores irrespective of the nature of the test. Interestingly, none of the personality variables was a significant predictor of the dependent variables in the three regressions. The "big three" personality factors were not significant predictors of the ability scores - which is in accordance with a good deal of previous research (Furnham et al., 1998b).
These results do not necessarily imply that dissimulators are necessarily unintelligent. It does mean that (personality) test dissimulators do, however, tend to emphasize their positive qualities. However, there is also a sizeable literature on various groups like psychiatric patients and insurance claimants who do the opposite, namely fake bad.
What is perhaps most interesting is that a lie score on a personality test consistently predicts scores on two ability tests, even though only 5 to 15% of the variance can be accounted for. This implies that dissimulation may be cross-test and perhaps cross-situationally consistent.
These results cannot, and do not, offer an explanation for the process underlying the associations between the variables examined, though various speculations are possible. People may come to be dissimulators about their intelligence and abilities over time because of self-doubt over their actual abilities. These beliefs may - or may not - be based on actual intelligence, but they lead some individuals, particularly in a job selection interview situation, to dissimulate on all measures, on tests of both power and preference. However, it should not be possible to dissimulate on tests of power except by under-performing. What it does suggest, however, is that self-report concerning intelligence and ability may be particularly unreliable - particularly in a job interview. Further, in that meta reviewers have shown that psychometric intelligence is the best predictor of success at work it appears important that it is measured by objective test rather than by any self-report method (Cook, 1998).
Baddeley, A. (1968). A three-minute reasoning test based on grammatical transformations. Psychometric Science, 10, 341-342.
Brebner, J., & Stough, C. (1995). Theoretical and empirical relationships between personality and intelligence. In D. Saklofske & M. Zeidner (Eds.), International handbook of personality and intelligence (pp. 321-347). London: Plenum Press.
Cook, M. (1998). Personnel selection. Chichester: Wiley.
Costa, P., & McCrae, R. (1995). Primary traits of Eysenck's P-E-N system. Journal of Personality and Social Psychology, 69, 308-317.
Eysenck, H. (1971). Relationship between intelligence and personality. Perceptual and Motor Skills, 32, 637-638.
Eysenck, H., Barrett, P., Wilson, G., & Jackson, C. (1992). Primary trait measurement of the 21 components of the P-E-N system. European Journal of Psychological Assessment, 8, 109-117. Furnham, A. (1986). Response bias, social desirability and dissimulation. Personality and Individual Differences, 7, 385-400.
Farnham, A., Forde, L., & Cotter, T. (1998a). Personality scores and test taking style. Personality and Individual Differences, 24, 19-23.
Farnham, A., Forde, L., & Cotter, T. (1998b). Personality and intelligence. Personality and Individual Differences, 24, 187-192.
Farnham, A., Gunter, B., & Peterson, E. (1994). Television distraction and the performance of introverts and extraverts. Applied Cognitive Psychology, 8, 705-711.
Hammerton, M. (1969). Interference between low information verbal output and a cognitive task. Nature, 222, 196-197.
Iverson, I. (2001). Can malingering be identified with the judgment of line orientation test? Applied Neuropsychology, 8, 167-173.
Iverson, I., Slick, D., & Franzen, D. (2000). Evaluation of WMS-R malingering index in a non-litigating clinical sample. Journal of Clinical and Experimental Neuropsychology, 22, 191-197.
Iverson, I., Franzen, D., & McCracken, L. (1992). Evaluation of an objective assessment technique for the detection of malingered memory deficits. Law and Human Behaviour, 15, 667-676.
Jackson, C., Fumham, A., Forde, L., & Cotter, T. (2000). The structure of the Eysenck Personality Profiler. British Journal of Psychology, 91, 223-239.
Matthews, G., & Dom, L. (1995). Cognitive and attentional processes in personality and intelligence. In D. Saklofske & M. Zeidler (Eds.), International handbook of personality and intelligence (pp. 367-396). London: Plenum Press.
Paulus, D. (1991). Measurement and control of response bias. In J. Robinson, P Shaver & L. Wrightsman (Eds.). Measurement of Personality and Social Psychological Attitudes. London: Academic Press. Vol. 1, pp. 17-59.
Rawlings, D., & Camie, D. (1989). The interaction of EPQ extraversion and WAIS subtest performance under timed and untimed conditions. Personality and Individual Differences, 10, 453458.
Robinson, D. (1985). How personality relates to intelligence test performance: Implications for a theory of intelligence, ageing research and personality assessment. Personality and Individual Differences, 6, 203-316.
Saklofske, D., & Kostura, D. (1990). Extraversion-introversion. Personality and Individual Differences, 11, 547-551.
Slick, D., Hopp, G., Strauss, E., & Hunter, M. (1994). Detecting dissimulation. Journal of Clinical and Experimental Neuropsychology, 16, 472-481.
Wilson, G., & Jackson, C. (1994). The personality of physicists. Personality and Individual Differences, 16, 187-189.
Wonderlic, E. (1992). Wonderlic Personnel Test. IL: Libertyville.
ADRIAN FURNHAM University College London, United Kingdom
Professor Adrian Furnham, Department of Psychology, University College London, United Kingdom.
Acknowledgement is due to reviewers including Dr. Donald H. Saklofske, Department of Educational Psychology, University of Saskatchewan, Saskatoon, Canada.
Please address correspondence and reprint requests to: Professor Adrian Furnham, Department of Psychology, University College London, 26 Bedford Way, London WCI OAP, UK. Phone: 020 76795395: Fax: 020 7436-4276; Email: