Personality & Individual Differences Vol. 4, No. 3, pp. 351-353, 1983.
Psychopathy, anxiety and malingering
JOHN J. RAY
School of Sociology, University of New South Wales, P.O. Box l, Kensington, N.S.W., Australia 2033
(Received 23 August 1982)
Summary --- Psychopathy is commonly measured as a low score on an Anxiety scale or as a high score on the MMPI Pd scale. The Pd scale, the Spielberger et al. Trait Anxiety scale and a short Achievement Motivation scale were administered to a random postal sample of 115 Australians. It was found that there was a high positive correlation between the Pd and Anxiety scales. This strongly supports Eysenck's theory that psychopaths are neurotic extraverts. A possible alternative explanation of the findings, however, is that the Pd scale may measure not psychopathy but malingering.
Although some cynicism concerning the MMPI seems to have become rather prevalent in recent years, Golden (1979, p. 57) is of the view that it still is the most widely-used psychological test in clinical applications. Although it is now quite an old test and one that might very reasonably be expected to have become dated, it is still widely used in research into psychopathy and Hare (1970, p. 15) reports that it still discriminates quite well between not only psychopaths and normals but also between psychopaths and other deviants. The scale within the MMPI originally designed to pick out psychopaths is the Pd scale but psychopathy diagnoses have been found to be strengthened if high Pd scores are found to be combined with high Mania scores.
There is of course considerable disagreement over who should be classified as psychopathic (or sociopathic) among both psychiatrists and psychologists. What limited agreement does exist among psychologists, however, seems to focus either on the MMPI or on Anxiety scales as measurement methods (Hare, 1970; Hare and Schalling, 1978). A psychopath is picked out either as a high Pd scorer or as a low scorer on any one of a variety of Anxiety scales (Hare, 1970, p. 17). One should be able to infer, then, that these two measurements would show at least some negative (inverse) correlation.
In making this inference, however, one must be careful to specify that one is concerned with primary, rather than secondary psychopaths. Hare (1970) points out that in addition to the "true", psychopath, there are also some antisocial individuals who, although aggressive and otherwise somewhat like psychopaths do nonetheless at the same time have strong feelings of guilt and anxiety. He refers to such people as "secondary" psychopaths and argues that they are, by and large, irrelevant to the study of psychopathy as such since their behaviour results from apparently quite ordinary stresses and conflicts -- as distinct from the true or primary psychopath whose motivations generally defy conventional analysis. Since human behaviour generally does very strongly tend towards being multicausal, the possibility that different causes could be producing similar effects is not an unreasonable one. At the risk of enraging behaviourists, one might say that one can act like a psychopath without necessarily being one. It may be worth noting that Blackburn (1975) used empirical classificatory procedures (cluster analysis) on a group of criminal offenders which produced categories very similar to those used by Hare. Blackburn's Types 1 and 2 might almost be a recapitulation of the primary/secondary distinction.
Given, then, that low anxiety is almost definitionally an attribute of the true psychopath, the work of Eysenck comes as something of a surprise. Eysenck (1964) claims that psychopaths are neurotic extraverts. Since neuroticism in his usage seems very largely equivalent to the concept of chronic anxiety, there clearly are diametrically-opposite conclusions to be drawn from his work and the work of other writers such as Hare (1970) and Cleckley (1976). Is the psychopath anxious or not?
One possible resolution of the difficulty is to say that Eysenck's work is concerned solely with "secondary psychopaths". At least in his earlier work, however, Eysenck makes no such qualification. He appears to be discussing psychopathy in general. Another possible resolution is to question the adequacy of Eysenck's measuring instruments. Hare (1970) points out that Eysenck's N scale generally seems to give mean scores among psychopaths which are very similar to the general population norm. On this evidence, one might say that the N scale is sensitive to neither primary nor secondary psychopathy. As the N scale does however seem in other research to be a reasonably good measure of anxiety, such a failure only serves to deepen the mystery about the relationship between psychopathy and anxiety. A possible way out of the mystery might be if it could be shown that the research Hare cites used a group of psychopaths who were equally mixed between "primary" and "secondary" types. Eysenck, himself seems largely to have lost interest in his earlier hypothesis-preferring to identify psychopaths in his more recent writings as high scorers on his new P scale (Eysenck and Eysenck, 1978).
Eysenck's hypothesis is however not without support from other writers. In their study of the factor structure of the MMPI, Kassebaum, Couch and Slater (1959) not only found two primary factors very similar to Eysenck's N and E but they also found that the Pd scale loaded 0.71 on the N factor. As the subjects in their study were however 160 Harvard freshmen, the generalizability of their results must remain speculative. In some contrast to the Kassebaum study, another study of the Pd scale itself by Monroe, Miller and Lyle (1964) found six factors -- none of which were labelled as measuring anxiety, neuroticism or similar constructs. Fenz, Young and Fenz (1974) also found that it was only secondary psychopaths who were neurotic.
In summary, then, it seems that Eysenck's position is a minority one -- though not without some outside support. The general conclusion emerging from the literature would seem to be that primary psychopaths are not anxious and that Pd score is probably the most widely-used measure of primary psychopathy that there is. Pd score should therefore correlate negatively with a measure of anxiety. The present study was designed to test this prediction on a sample offering somewhat more prospects of generalizability than that used by Kassebaum et al. (1959). Another much more secondary aim of the present work was to explore the relationship between achievement motivation and psychopathy. Whatever else they are, psychopaths are not notable for their foresight about the future and the sort of materialistic ambition that seems commonly to be measured in achievement-motivation studies would seem very strongly to require long-term planning and foresight. The hypothesis that seems indicated, then, is that psychopaths will score low on measures of achievement motivation.
The most well-known and widely-used scale of anxiety at present appeared to be the Spielberger, Gorsuch and Lushene (1970) "Trait" Anxiety scale. Lykken, Tellegen and Katzenmeyer (1973) have produced an Anxiety scale specifically designed for use with psychopaths (the APQ) but a more generally-usable scale seemed most appropriate for use on the present occasion. To the Spielberger and Pd scale was added a third scale-the Ray (1979, 1980) AO scale (to measure achievement motivation) in its 14-item short form. This is a well-validated scale that does stress in its items the sort of systematic ambition of a planned and lifelong kind that is hypothesized to be low in psychopaths.
All three scales were combined into a questionnaire and mailed out to 400 addresses chosen at random from the electoral rolls (voter registration lists) of the Australian state of New South Wales. Voter registration in Australia at the time was compulsory not only for all Australian citizens over the age of 18 years but also for many non-citizen immigrants. Because of this unusual system, Australian electoral rolls do therefore provide an unusually comprehensive sampling frame. The major omissions would be transients and citizens of countries outside the British Commonwealth. Previous one-wave surveys of the present type by the same author had shown a response rate of generally around 25% so a final n of about 100 was envisaged. Although such an n might seem small for some purposes, the present purpose of obtaining a sample which would give a general idea of the relationship between the three scales did not seem to require anything more elaborate.
The use of a general population sample rather than a clinical group as Ss is of course unusual for the MMPI but it was felt that only thus could maximally-generalizable results be obtained. Cleckley (1976) and many others (see Ray and Ray, 1982) are of the view that psychopathic tendencies are quite common in the population at large so the present enquiry is simply broader than would be an enquiry based on the responses of a small, institutionalized minority. We are asking then: are psychopathic tendencies and anxiety opposed in general?
The number of questionnaires finally returned was 115. These then formed the sample for the study. Previous use of the Pd scale in postal surveys had revealed (Ray and Ray, 1982) that several of its items did not function well in such a context. A total of 18 out of 50 items showed poor correlations with the total score on the scale. Since the scale formed of the remaining 32 items correlated 0.937 with the 50-item scale, however, it seemed that whatever was true of the one version of the scale must also be true of the other. It was therefore the 32-item form of the scale that was scored on the present occasion. The reliabilities (alpha) observed for the three scales were: Pd, 0.86; AO, 0.78; Anxiety, 0.91. All three scales did then by this criterion seem to be functioning as intended.
The correlation between the Pd and Anxiety scales was 0.676 -- a very strong relationship as such relationships go but in the direction opposite to that predicted. The AO scale correlated as predicted with the Pd scale (r = -0.179) but in the direction opposite to that predicted with the Anxiety scale (r = -0.142). Only the former of the two is however significant at the 0.05 level.
Since the result obtained was so very different from that expected, it seemed worth investigating whether or not the same relationship would obtain if only high scorers on the Pd scale were used. Constricting the range of scores of a variable does of course tend to restrict the range of correlations it can show but the interest was to see whether the unexpected direction of the correlation would be preserved. A subsample was therefore selected which consisted of all those who scored above the Pd scale mean. This consisted of 54 people. The correlation between the Pd and Anxiety scales on this subsample was 0.631.
At least at first sight, the present results do seem to support very strongly the account of psychopathy given by Eysenck. We take the most widely-used measure of primary psychopathy and find that it correlates strongly positively with a widely-used measure of anxiety. The sample used also gives this finding considerable prospect of generalizability.
No doubt, however, the immediate response of Hare and others would be to question the validity of the Pd scale -- even though such questioning must also represent an attack on much prior work in the field. Given the fact that the Pd scale was "purpose-built" for the measurement of psychopathy and given its empirical method of construction (selecting item that discriminate known groups), however, the presumption of validity must at least initially lie with the Pd scale. This impression is reinforced by the concurrent validation furnished by the AO scale on the present occasion.
Nonetheless, the fact that the present finding runs counter to so much in the way of previous findings (Hare, 1970; Cleckley, 1976) must make us careful to seek alternative explanations of the present results. In this vein, an obvious objection might be that the present study was not in fact a study of psychopaths as a group. This should however probably not be totally conceded. The separate analysis of the results among the high-scoring (on the Pd scale) subsample should provide at least some check on how "real" psychopaths might score. Even this analysis might be seen as pointless, however if it is argued that psychopathy is a solely clinical entity with no subclinical manifestations. Such an argument would however represent a form of dogmatism that flies in the face of most current psychological-measurement procedures.
A more fruitful alternative explanation of the present results and one more in line with current trends in social psychological explanation might be to look at the demand characteristics of the situation in which the original criterion group of clinical psychopaths answered the MMPI. They were of course an institutionalized group and it is well documented (Cleckley, 1976) that institutionalized psychopaths commonly use the mental-health system to escape the consequences their deeds. By making a plea of mental illness they often succeed in avoiding or getting out of prison. It might therefore be true that the criterion group of psychopaths used in the original study saw it as being in their interests to feign a degree of personal distress that they did not at all feel. The Pd scale might, in other words, be little more than a malingering scale. It might even be generally true that institutionalized psychopaths get high scores on it but they do so only because such people are generally malingerers in such a situation. It might be interesting to see what scores they get when they are in the phase of trying to get released from hospitalization. Certainly when one looks at the items of the Pd scale it is very easy to see why they correlate so highly with anxiety. They cover in fact a very full range of claims of personal distress. It might also be significant that the mean score on the scale was very low relative to the possible range of scores. If the scale midpoint is taken as the number of items multiplied by the midpoint score for each item, we get a midpoint score of 96 (32 x 3). The actual mean was a much lower 72.81 (SD 14.77). In other words, the claims of distress that characterize the scale were rated by most respondents as "False" for them.
In conclusion, it seems clear that much more sophisticated precautions against faking will have to be employed in future than have been employed in the past if any advance in our systematic knowledge about psychopaths is to take place at all.
Blackburn R. (1975) An empirical classification of psychopathic personality. Br. J. Psychiat. 127, 456-460.
Cleckley H. (1976) The Mask of Sanity, 5th edn. Mosby, St Louis.
Eysenck H. J. (1959) Manual of the Maudsley Personality Inventory. Univ. London Press.
Eysenck H. J. (1964) Crime and Personality. Methuen, London.
Eysenck H. J. and Eysenck S. B. G. (1978) Psychopathy, personality and genetics. Psychopathic Behaviour: Approaches to Research, Chap. 12 (Edited by Hare R. D. and Schalling D.). Wiley, Chichester, England.
Fenz W. D., Young M. J. and Fenz H. G. (1974) Differences in the modulation of cardiac activity between psychopaths and normal controls. Psychosom. Med. 36(6), 488-502.
Golden C. J. (1979) Clinical Interpretation of Objective Psychological Tests. Grune & Stratton, New York.
Hare R. D. (1970) Psychopathy: Theory and Research. Wiley, New York.
Hare R. D. and Schalling D. (Eds) (1978) Psychopathic Behaviour: Approaches to Research. Wiley, Chichester, England.
Kassebaum G. G., Couch A. S. and Slater P. E. (1959) The factorial dimensions of the MMPI. J. consult. Psychol. 23, 226-236.
Lykken D. T., Tellegen A. and Katzenmeyer C. (1973) Manual for the Activity Preference Questionnaire (APQ) Report No. PR-73-4 from the research laboratories of the Department of Psychiatry, University of Minnesota.
Monroe J. J., Miller J. S. and Lyle W. H. (1964) The extension of psychopathic deviancy scales for the screening of addict patients. Educ. psychol. Measur. 14, 47-56.
Ray, J.J. (1979) A quick measure of achievement motivation -- validated in Australia and reliable in Britain and South Africa. Australian Psychologist 14, 337-344.
Ray, J.J. (1980) The comparative validity of Likert, projective and forced-choice indices of achievement motivation. J. Social Psychology, 111, 63-72.
Ray, J.J. & Ray, J.A.B. (1982) Some apparent advantages of sub-clinical psychopathy. Journal of Social Psychology, 117, 135-142.
Spielberger C. D., Gorsuch R. L. and Lushene R. E. (1970) Manual for the State-Trait Anxiety Inventory. Consulting Psychologists Press, Palo Alto, California.
Go to Index page for this site
Go to John Ray's "Tongue Tied" blog (Backup here or here)
Go to John Ray's "Dissecting Leftism" blog (Backup here or here)
Go to John Ray's "Australian Politics" blog (Backup here or here)
Go to John Ray's "Gun Watch" blog (Backup here or here)
Go to John Ray's "Education Watch" blog (Backup here or here)
Go to John Ray's "Socialized Medicine" blog (Backup here or here)
Go to John Ray's "Political Correctness Watch" blog (Backup here or here)
Go to John Ray's "Greenie Watch" blog (Backup here or here)
Go to John Ray's "Food & Health Skeptic" blog (Backup here)
Go to John Ray's "Leftists as Elitists" blog (Not now regularly updated -- Backup here)
Go to John Ray's "Marx & Engels in their own words" blog (Not now regularly updated -- Backup here)
Go to John Ray's "A scripture blog" (Not now regularly updated -- Backup here)
Go to John Ray's recipe blog (Not now regularly updated -- Backup here or here)
Go to John Ray's Main academic menu
Go to Menu of recent writings
Go to John Ray's basic home page
Go to John Ray's pictorial Home Page (Backup here)
Go to Selected pictures from John Ray's blogs (Backup here)
Go to Another picture page (Best with broadband)