Psychologica Belgica
1990, 30.2/3, 87-111
PERSONALITY ASSESSMENT IN VIOLENT OFFENDERS:
THE DEVELOPMENT OF THE
ANTISOCIAL PERSONALITY QUESTIONNAIRE
Ronald BLACKBURN
University of Liverpool, UK
The Antisocial Personality Questionnaire is a 125-item inventory that meas-
ures eight primary traits (self-control, self-esteem, avoidance, paranoid suspi-
cion, resentment, aggression, deviance, extraversion) and two higher-order
dimensions (impulsivity-aggression versus control, withdrawal versus socia-
bility). The questionnaire was developed for assessing dangerous mentally
disordered offenders, and four profile patterns of primary psychopath, secon-
dary psychopath, controlled personalities, and inhibited personalities have
been consistently identified in this population. The derivation of the APQ
from an earlier questionnaire is described, and evidence for its validity pre-
sented, Although developed with violent offenders, the APQ is likely to be of
use in discriminating within offender populations generally,
Research since the 1950s has regularly found that criminal samples are
distinguished by traits exemplifying a “weak superego” (Andrews &
Wormith, 1989). For example, a study of a New Zealand birth cohort at age
18 found that self-report personality profiles of both male and female delin-
quents were characterised by impulsivity, danger seeking, a rejection of tra-
ditional values, aggressive attitudes, feelings of alienation, and an adversa-
rial interpersonal attitude (Caspi, Moffitt, Silva, Stouthamer-Loeber,
Krueger & Schmutte, 1994). These findings were replicated in a younger
American sample. A similar profile also predicted a DSM-III-R (American
Psychiatric Association, 1987) diagnosis of antisocial personality disorder
three years later (Krueger, Caspi, Moffitt, Silva & McGee, 1996).
As Krueger et al. (1996) note, this trait pattern resembles the “classical”
concept of psychopathic personality proposed by Cleckley (1976), and
research with the Psychopathy Checklist-Revised (PCL-R: Hare, 1991,
1996) has shown that within offender populations, the more psychopathic
tend to have the most extensive criminal careers. The PCL-R also distin-
guishes the more violent offenders (e.g., Blackburn & Coid, 1998; Kosson,
Smith, & Newman, 1990), and is a significant predictor of both general and
violent recidivism (Salekin, Rogers, & Sewell, 1996).
1 am indebted to Diane Fawcett for her help with data collection and to Brian Thomas-Peter
for supplying data on the probation sample.
Correspondence conceming this article should be addressed to Ronald Blackburn,
Department of Clinical Psychology, University of Liverpool, Liverpool L69 3GK, UK.88 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
However, offenders are not homogeneous in personality, and this applies
particularly to those who commit violent crimes. For example, a recent study
of violent offenders identified three patterns of personality disorder associa-
ted with antisocial personality disorder and high PCL-R scores, and these
patterns were distinguished by both offence history and psychopathology
(Blackburn & Coid, 1999). Moreover, although aggressive traits may make
violence more likely, violent crimes depend on situational and temporary
emotional factors as well as other dispositions a person brings to a situation,
and are often committed by people who are not habitually aggressive. This
is recognised in the hypothesis that violent offenders can be divided into
overcontrolled and undercontrolled types (Megargee, 1966). For purposes of
understanding serious criminal acts, predicting recidivism, or intervention,
personality assessment therefore needs to take account of traits in addition
to those associated with psychopathy.
This paper describes the development of the Antisocial Personality
Questionnaire (APQ: Blackburn & Fawcett, 1996, 1999), a multi-trait, self-
report inventory which measures cognitive, affective and behavioural dispo-
sitions of relevance to offender populations generally, and violent offenders
in particular. During the past 15 years, several comprehensive multitrait self-
report measures of personality deviation have been developed as a result of
growing interest in personality disorders (e.g., Millon, 1983; Clark, Livesley,
Schroeder, & Irish, 1996). However, few inventories have been developed
specifically for use with offenders. The APQ provides a relatively brief and
reasonably comprehensive assessment of deviant traits and has the advanta-
ge of having been validated in a population of violent offenders. The varia-
bles assessed do not represent a single theory of criminality or personality,
but the questionnaire samples beliefs about the self and others, aspects of
emotional control, and interpersonal problem behaviours which have figured
prominently in theory and research on antisocial behaviour (Andrews &
Wormith, 1989; Blackburn, 1993; Wilson & Herrnstein, 1985).
The APQ is derived from an earlier questionnaire, the Special Hospitals
Assessment of Personality and Socialisation (SHAPS: Blackburn, 1982).
The APQ measures broadly the same variables but with fewer items, and
replaces the SHAPS. Research with the SHAPS is therefore described first.
The Special Hospitals Assessment of Personality and Socialisation
The SHAPS originated from research and clinical work with adult male
mentally disordered offenders detained in English maximum security hospi-
tals (special hospitals) because of their presumed dangerousness. Some of
these offenders are classified under the Mental Health Act as suffering fromTHE ANTISOCIAL PERSONALITY QUESTIONNAIRE 89
psychopathic disorder (‘...a persistent disorder or disability of mind...
which results in abnormally aggressive or seriously irresponsible conduct on
the part of the person concemed”), and early work with the SHAPS was par-
ticularly concerned with determining the relationship of this legal category
to the clinical concept of psychopathic personality (Blackburn, 1975a).
However, personality deviation is often significant in the antisocial beha-
viour of offenders identified as mentally ill, who form the majority of
patients in special hospitals. The SHAPS has therefore been used for person-
ality assessment with mentally disordered offenders generally. It has also
been used in research with prisoners in England (Clark, 1996; Needs, 1986),
Canada (Kroner, 1997; Kuriychuk, 1990), and the United States (Como,
1977, Widom, 1978), and with community samples of delinquents (Gillham,
1978; Renwick & Emler, 1991), and sex offenders (Beckett, Beech, Fisher,
& Fordham, 1994). These studies indicate that the SHAPS scales discrimi-
nate within offender populations generally,
SHAPS scales
During the 1970s, the MMPI was the most frequently used personality
questionnaire in the special hospitals, but despite its rich sampling of psy-
chological problems, this inventory has several limitations when applied to
mentally disordered offenders. In particular, it fails to separate personality
traits from symptoms of illness (Foulds, 1971), and traits important in the
assessment of antisocial behaviour, such as aggression or hostility, are not
well represented. There is also considerable redundancy in the item pool.
The Neuroticism and Extraversion dimensions pervade the MMPI, and most
items are heavily weighted by Neuroticism or negative affectivity
(Kassebaum, Couch & Slater, 1959; McCrae & Costa, 1986).
The SHAPS was formed by selecting potentially useful scales from the
MMPI. To these were added measures of hostility and aggression derived
from factor analysis of the Buss-Durkee Hostility Inventory (Buss, 1961)
and the Psychopathy scale of Peterson, Quay & Cameron (1959). The
213 items were rephrased as questions requiring a YES-NO response to faci-
litate administration to patients of low intelligence or limited literacy. The
SHAPS contains 10 scales (Lie, Anxiety, Extraversion, Hostility, Shyness,
Depression, Tension, Psychopathic Deviate, Impulsivity, Aggression).
Anxiety (A: Welsh, 1956) and Extraversion (EX: Giedt & Downing, 1961)
are markers for the Neuroticism and Introversion-Extraversion dimensions,
respectively (Corah, 1964). Lie (L) and Psychopathic Deviate (PD) are stan-
dard MMPI scales, and the Shyness (1), Depression (D), and Tension (T)
scales were drawn from a cluster analysis of MMPI items (Stein, 1968).
Impulsivity (IM), Hostility (HO), and Aggression (AG) were developed wit-90 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
hin the special hospitals (Blackbur, 197 1a, 1972, 1974, 1975b).
At the time the SHAPS was developed, there was agreement that the
higher order dimensions of Neuroticism and Extraversion accounted for
much of the variation in personality, but little consensus beyond that. These
dimensions are represented in the SHAPS by A and Ex. Factor analyses,
however, indicate that the higher order factors in the questionnaire do not
align directly with these dimensions
Anearly study attempted to clarify the relationship of aggression and hos-
tility to Neuroticism and Extraversion by means of a factor analysis of sever-
al personality measures (Blackburn, 1972). Three oblique factors of aggres-
sion, hostility-neuroticism, and extraversion emerged. The aggression factor
was defined by scales of overt aggressiveness, acting out, resentment, and
impulsivity, and negatively by Megargee’s Overcontrol scale. This factor
correlated positively with hostility-neuroticism but was independent of
extraversion. It may have some relationship to what is now identified as the
Agreeableness (versus antagonism) dimension of the Big Five (Costa &
McCrae, 1992), but these results indicated two higher-order factors.
Subsequent analyses (Blackburn, 1979a, 1986) consistently show that
these two factors account for most of the variance in the 10 SHAPS scales.
The factors are Impulsive Aggression/Psychopathy versus Control (IM, AG,
and HO vs L) and Withdrawal versus Sociability (I vs EX). This is not an
artefact of sampling, because the same rotated factors emerge in nonoffen-
der volunteers as in mentally disordered offenders (Table 1). Blackburn
Table 1
Rotated Factor Loadings of SHAPS Scales in Patient and Normal Samples
Patients (N = 499) Normals (N = 238)
Factors Factors
SHAPS scale 1 1 1 nw
Lie -4 -.05 -55 -03
Anxiety 68 65 34 76
Extraversion 35 -84 48 ~16
Hostility 76 28 2 44
Shyness 2 86 -01 89
Depression 58 2 a2 B
Tension 66 56 53 63
Psychopathic Deviate 61 39 By 29
Impulsivity 92 -.06 2 -.05
Aggression 86 \4 86 i
% variance 4s 2» 38 32
Note. SHAPS = Special Hospitals Assessment of Personality and Socialisation,THE ANTISOCIAL PERSONALITY QUESTIONNAIRE 91
(1996) found that these factors correspond to the two largest factors in the
personality disorder scales of the Millon Clinical Multiaxial Inventory
(MCMI-1: Millon, 1983). The factor equivalent to impulsivity in the MCMI-
1 is defined by the passive-aggressive, paranoid, borderline, and antisocial
scales versus the compulsive scale, while the factor parallel to withdrawal is
defined by schizoid, avoidant, and schizotypal versus histrionic and narcis-
sistic scales. In an analysis of interview ratings of DSM-III personality dis-
orders in violent offenders, Blackburn and Coid (1998) identified concep-
tually similar factors of hostile impulsivity (paranoid, histrionic, narcissistic,
antisocial, borderline, and passive-aggressive disorders) and detachment
(schizoid, schizotypal versus histrionic). The first factor correlated highly
with Hare’s PCL-R and with criminality. Although the convergence of
questionnaire and interview measures of personality disorder is not general-
ly very close (Zimmerman, 1994), the similarities of these factors across dif-
ferent measures is striking.
The SHAPS factors therefore represent general dimensions that subsume
many of the deviant traits of personality disorder, and factor scales B
(Belligerence) and W (Withdrawal) were constructed to measure these
dimensions (Blackburn, 1987). The factors represent 45° rotations in the
two-dimensional space of Neuroticism and Extraversion (Kassebaum et al.,
1959). They therefore seem equivalent to Gray’s impulsivity and anxiety
dimensions (Gray, 1987), which are held to be manifestations of the
Behavioural Activation System (BAS) and Behavioural Inhibition System
(BIS), respectively. In view of the contribution of aggression to the first fac-
tor, it is possible that the SHAPS factors are higher order dimensions incor-
porating Agreeableness as well as Neuroticism and Extraversion, but rela-
tionships with the Big Five dimensions remain to be tested.
Validity of SHAPS Scales
The valdity of the scales as personality measures has been supported in
several studies within offender samples. Scales loading the first factor (IM,
AG) correlate significantly with ratings of aggression (Blackburn, 1974,
1975b; Blackburn & Maybury, 1985; Davis, 1974), anger proneness
(Blackburn & Lee-Evans, 1985), interpersonal style (Willner & Blackburn,
1988), self-reported delinquency (Renwick & Emler, 1991), autonomic and
electrocortical activity (Blackburn, 1975b, 1979b; Howard, Fenton, &
Fenwick, 1984), serotonergic function (Dolan, 1996), and moral reasoning
(O'Kane, Fawcett, & Blackburn, 1996). Scales related to the second factor
(EX, 1) correlate with social skills (Renwick & Emler, 1991) and interper-
sonal style (Willner & Blackburn, 1988).92 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
Table 2
Corretations of SHAPS B and W Scales with the PCL-R in Three Samples of Male Offenders
PCL-R = PCL-R PCL-R
Sample N Total Factor 1 Factor 2
Kuriychuk (1990) 60 prison inmates and staff
B agit 31 36"
w 7 224 -09
O'Kane et al. (1996) 39 mentally disordered offenders
B ae 9 30"
w 15 15 20
Kroner (1997) 97 prison inmates
B goer aye agnor
Ww 2 16 25
Note. SHAPS = Special Hospitals Assessment of Personality and Socialisation; B = Belligerence: W =
Withdrawal; PCL-R = Psychopathy Checklist-Revised.
oF = p<. =p <0l;*=p<.05
All scales have correlates with the personality disorder scales of the
MCMI-I (Blackburn, 1996), and the SHAPS factors as measured by B and
W also correlate with the main factors of the MCMI-1. B correlates .77 with
the MCMI-1 passive-aggressive-versus compulsive factor, while W correla-
tes .80 with the MCMI-1 schizoid versus histrionic factor (N = 133).
Correlations of the B and W scales with Hare’s PCL-R have also been obtai-
ned in samples of Canadian prisoners (Kroner, 1997; Kuriychuk, 1990) and
mentally disordered offenders in England (O’Kane et al., 1996). As can be
seen in Table 2, B correlates moderately but consistently with PCL-R total
score and scores on PCL-R Factor 2 (social deviance). The correlation of B
with PCL-R Factor 1 (callous, remorseless style) is less consistent.
Correlations of W with the PCL-R are small, and reach a significant level in
only one sample.
These studies support the construct validity of several of the SHAPS sca-
les. The scales have also been found to have predictive utility. Clark (1996)
reports a preliminary 5-6 year follow-up of 117 life sentenced prisoners that
suggests that SHAPS scales may predict progress through the penal system.
Prisoners transferred to psychiatric establishments had higher scores on A,
D, and PD, and lower scores on EX. Prisoners who survived the system more
adaptively as shown by transfer to less secure prisons tended to score low on
D and AG. Black and Spinks (1985) included some SHAPS scales ina 5 year
follow-up of mentally disordered offenders discharged from a maximum
security hospital. Those surviving most successfully in the community had
low scores on PD and IM, while EX and IM were significantly related toTHE ANTISOCIAL PERSONALITY QUESTIONNAIRE 93
further court appearances and violent recidivism. Como (1977) also found a
significant association of IM with general recidivism in an American prison
sample.
An Empirical Typology of Offenders
SHAPS scales permit descriptive statements about an offender's salient
personality characteristics, but certain patierns of scores are relatively com-
mon, Common profiles are likely to represent naturally occurring personali-
ty types that are significant in understanding antisocial behaviour. Several
studies have therefore been carried out with the aim of reducing the hete-
rogeneity of the mentally disordered offender population by means of clus-
ter analysis.
To test Megargee’s hypothesis that violent offenders can be divided into
overcontrolled and undercontrolled types (Megargee, 1966), Blackburn
(1971b) carried out a cluster analysis of MMPI profiles of mentally disorde-
red offenders who had committed homicide. The results indicated two
undercontrolled and two overcontrolled profile types. One undercontrolled
type was defined by the well known “49” profile associated in the clinical
literature with psychopathic personality (i.e., combined elevations on the PD
and Hypomania scales), the other by a highly deviant profile which includes
abnormal scores on most other clinical scales, Of the overcontrolled groups,
one was defined by a defensive “hypernormal” pattern, the other mainly by
marked social introversion and dysphoria.
This fourfold typology is consistently reproduced by cluster analyses of
the SHAPS scales or factor scores. The same four profiles were distinguis-
hed among patients in the legal category of psychopathic disorder
(Blackburn, 1975a), but they also constitute the main patterns of personality
deviation among mentally disordered offenders as a whole (Blackburn,
1986). The four classes are: (1) Primary Psychopaths (P; impulsive, aggres-
sive, hostile, extraverted, self-confident, low to average anxiety); (2)
Secondary Psychopaths (S; hostile, impulsive, aggressive, socially anxious,
withdrawn, moody, low in self esteem): (3) Controlled (C; defensive, con-
trolled, sociable, nonanxious); (4) Inhibited (1; shy, withdrawn, controlled,
moderately anxious, low self esteem). The four groups represent combina-
tions of extremes on the two SHAPS factor dimensions. P and S score
towards the impulsive-aggressive extreme of the first factor, but are distin-
guished by the withdrawal-sociability dimension, P being socially uninhibi-
ted, § withdrawn and anxious. The C and I groups score at a low level on the
impulsivity factor, but are also differentiated by withdrawal-sociability.
The classification has been replicated in research in English prisons on
“normal” murderers (McGurk, 1978), violent offenders (Henderson, 1982),94 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
and also unselected prisoners (McGurk & McGurk, 1979). Widom (1978)
identified similar types in American female prisoners. The typology is the-
refore robust and appears to represent the main personality types identifia-
ble through self-report measures in prison and forensic psychiatric popula-
tions. These groups were also recovered from cluster analysis of MCMI-1
personality disorder scales (Blackburn, 1996). The same types seem likely
to emerge among antisocial populations from any comprehensive set of self-
report measures of personality deviation because of the common influence
of the two Big Five dimensions of Neuroticism and Extraversion. However,
the sources of differentiation do not lie directly in these factors, but rather in
their interaction in the impulsive-aggression and withdrawal dimensions,
A number of differences have been found between the “psychopathic”
groups and C and I, but also between P and S. Although the groups are repre-
sented throughout the mentally disordered offender population, P and S are
more predominant in the legal category of psychopathic disorder than among
the mentally ill (Blackburn, 1986, 1996). Blackburn (1975a) found that P
and S had earlier criminal careers than C and 1, although P had more con-
victions for violence. Similar findings on differences between the four
groups in criminal background were obtained by Pollock (1999) in a study
of homicides in Northern Ireland. He also found that the homicide offences
of primary psychopaths were more likely to be instrumental or goal-directed
and that their victims were more commonly strangers. The offences of C and
I offenders, on the other hand, were primarily reactive violence directed
against acquaintances.
P and S also score higher on the aggressive or antisocial personality dis-
order scale of the MCMI-1 (Blackburn, 1996). However, where P are also
narcissistic and histrionic, S are passive-aggressive, avoidant, schizoid,
dependent, and paranoid. The I group is also schizoid, avoidant, schizotypal,
and passive-aggressive, but differs from S in scoring low on the antisocial
scale. Controlled patients score highest on the compulsive disorder scale, but
show the least signs of personality disorder more generally.
Secondary psychopaths, who are characterised by the most deviant
MMPI and SHAPS profiles, are also typically most deviant in other respects
The EEG abnormalities claimed in the older clinical literature to be common
among psychopaths are most likely to be found in this group (Blackburn,
1979b; Howard, 1984), and they are also the least autonomically aroused
(Blackbum, 1979b). In a delinquent sample, Gillham (1978) also found that
S reported the least vivid emotional imagery. Primary psychopaths differ
from S in having higher levels of cortical and autonomic arousal (Blackburn,
1979b), and score highest on Zuckerman’s Sensation Seeking Scale
(Blackbum, 1978). The controlled group shows little distinctive that would
not be anticipated from their denial of strong emotional reactions or social-THE ANTISOCIAL PERSONALITY QUESTIONNAIRE 95
ly improper behavior, but the inhibited group includes prisoners who are the
least socially skilled (Henderson, 1982). Members of this group are also
more likely to be identified as mentally ill (Blackburn, 1996).
P and S describe themselves as more dominant in both threatening and
friendly interactions, but differences between the psychopathic and nonpsy-
chopathic groups are more apparent in threatening situations (Willner &
Blackburn, 1988). On the other hand, S describe the most intense anger in
response to verbal or physical threat (Blackburn & Lee-Evans, 1985). A
study in an American prison also found that secondary psychopaths who are
low in intelligence tend to be more violent than other inmates within the
prison setting (Heilbrun &Heilbrun, 1985).
These findings overall support the validity of this empirically derived
classification in discriminating classes of personality deviation among offen-
ders, The first group has been described as primary psychopaths because its
members are distinguished by a pattern of traits that closely approximate the
characteristics held to define the psychopath by Cleckley. In particular, their
hostile alienation from others, impulsivity, aggression, and a relative absen-
ce of anxiety or social inhibition seem consistent with this concept. Findings
that this group is distinguished by narcissism, sensation seeking, interper-
sonal dominance, and violent criminality strengthen this interpretation. The
secondary psychopaths share some of these traits, but differ in showing
extreme social anxiety and traits of schizoid, avoidant and passive-aggres-
sive personality disorders. The differences between the two groups are in
accord with the distinction made by Lykken (1957), who proposed that
Cleckley psychopaths could be divided according to high and low trait
anxiety.
The primary-secondary distinction has not been favoured in recent
research on psychopathy because it implies that there are two kinds of “psy-
chopath”, although Porter (1996) suggests that the distinction may be etio-
logically significant. Comparisons of the SHAPS types on Hare's PCL-R
suggest that although primary psychopaths are more likely to meet PCL-R
criteria of psychopathy, secondary psychopaths do not. Kuriychuk (1990)
found that 57% of primary psychopaths scored above the recommended
PCL-R cutoff of 30, compared with 15% of secondary psychopaths and 8%
of nonpsychopaths (controlled plus inhibited). Corresponding figures obtai-
ned by Blackburn (1998) were 45%, 14% and 8%. The label “secondary psy-
chopath” may therefore be misleading. Nevertheless, the group it refers to is
very real, and whatever the appropriate term, this is a clinically distinct and
deviant group which seems to be prevalent among offenders.96 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
The Antisocial Personality Questionnaire
Most SHAPS scales were developed through multivariate analysis, but
they are drawn from different studies employing different samples. Some
redundancy was built into the inventory, and it is likely that essential infor-
mation can be extracted from fewer variables. For example, correlations
between the A and D scales and between IM and Ag are typically in the
region of 0.8. These scale pairs have several items in common, and it is
unclear whether the correlations are artefacts of item overlap or reflect com-
mon underlying variables. In developing the APQ, the aim was to identify
the main primary or correlated traits in the SHAPS item pool and to con-
struct scales to measure these traits with minimal item overlap. Further
details of the APQ scale development can be found in a recent paper
(Blackburn & Fawcett, 1999).
The strategy adopted for scale development was a factor analysis of all
213 SHAPS items in a sample made up of 499 male mentally disordered
offenders and 238 volunteer “normal” males. The patient group consists of
366 patients from an earlier study, and a further sample of 133 for whom
additional data were obtained (validation sample). A third of the patients fell
in the legal category of psychopathic disorder, the remainder being legally
classified as mentally ill or mentally impaired (learning disabled). Their
mean age was 31.84 (SD = 10.41). The normal volunteer group consists
mainly of hospital and university staff, and had a mean age of 31.63 (SD =
9.90).
Principal components analyses were carried out within the patient and
normal samples separately and within the total sample of 737. The scree test
indicated 6 to 9 factors in the two subsamples and the total sample,and exa-
mination of varimax rotations within this range suggested an 8-factor solu-
tion. Items with highest loadings on each factor were subject to item analy-
ses which aimed to yield brief homogeneous scales (maximum 20 items)
with reliabilities (alpha) of at least .75. Maximum item overlap between sca-
les was kept to three items, and no item contributed to more than two scales.
The analyses resulted in 8 scales measured by 125 items, scale lengths
varying from 16 to 20 items.
In addition to the 8 primary factor scales, two scales were formed to
measure the higher order factors pervading the SHAPS and which are also
reproduced by the APQ scales (see below). Criteria for scale membership
were a high correlation (> 0.3) of an item with APQ factor scores on one fac-
tor and a low correlation (< 0.2) with scores on the other in both patient and
normal samples. This resulted in a 33-item scale measuring the impulsivity-
aggression factor (Impulsivity: 1) and a 24 item-scale measuring the with-
drawal factor (Withdrawal: W).THE ANTISOCIAL PERSONALITY QUESTIONNAIRE 97
APQ Scales
Self-Control (SC) (20 items). Items are drawn from the SHAPS
Impulsivity, Aggression, and Lie scales, and refer to denial of “improper”
behaviour (heavy drinking, swearing, excitement seeking), strong feelings
(anger, irritability), and negative interpersonal reactions (dislike, rudeness).
The scale correlates negatively with all other scales except Extraversion.
High scores may reflect a guarded, socially desirable presentation, but
external correlates suggest that they are open to more substantive interpreta-
tion in terms of rigid conformity, limited self-awareness or insight, or a cog-
nitive coping style that avoids information threatening to the self-concept.
High scores are common among chronic, stabilised psychotic patients. Low
scores reflect awareness and expression of undesirable impulses. Extreme
high and low scores are consistent with Megargee’s notions of overcontrol
and undercontrol, respectively.
Self-Esteem (SE) (18 items). Items are drawn mainly from the Anxiety
and Depression scales and refer to self-denigration, feelings of dysphoria
and helplessness, and worry. The scale represents the dimension of
Neuroticism or negative affectivity (Watson & Clark, 1984), but this is also
related to self-esteem, and the latter description is preferred here. Higher
scores reflect negative self-esteem and moodiness. Lower scores reflect
positive self-esteem, but extreme low scores may suggest lack of self-criti-
cism or denial.
Avoidance (AV) (16 items). This is essentially the Shyness (neurotic in-
troversion) scale of the SHAPS, items refering to difficulties in and avoidan-
ce of social interaction. High scorers are likely to be socially withdrawn and
unskilled, detached, and timid. Low scores are likely to be associated with
sociability and spontaneity.
Paranoid Suspicion (PA) (17 items). Items come mainly from the SHAPS
Hostility scale, and reflect marked suspicion and beliefs of personally direc-
ted malevolence on the part of others, particularly authority figures. Since
the scale correlates quite highly with avoidance and low self-esteem, high
scorers may not readily express their suspicions or grievances.
Resentment (RE) (19 items). This scale also contains items from the ori-
ginal Hostility scale, but includes some from other scales, notably
Psychopathic Deviate and Extraversion. Items reflect mistrust, but in the
context of a general cynicism about the trustworthiness and competence of
others. The scale correlates highly with PA, but where the latter suggests
vigilance for personal harm, RE seems to reflect a more general tendency to
externalise blame.
Aggression (AG) (20 items). Items come mainly from the original
Aggression scale, and describe anger arousal, impatience or loss of temper,98 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
and assaultative behaviour. High scores are likely to reflect a tendency to be
easily angered and to react with verbal and sometimes physical aggression.
Low scores suggest phlegmatism and compliance or restraint in the face of
provocation.
Deviance (DE) (20 items). Most items are drawn from the Psychopathic
Deviate scale. They reflect a deviant social history marked by rule-breaking
behaviour and delinquency and discordant intimate relationships, particular-
ly in the family. Low scores are likely to indicate conformity and socialisa-
tion. This is mainly a historical factor for which only limited change can be
expected, although negative perceptions of previous relationships may beco-
me less marked with successful clinical intervention.
Extraversion (EX) (20 items). Items come mainly from the SHAPS
Extraversion scale. While they emphasise sociability or gregariousness, they
also reflect an energetic involvement in and enjoyment of daily activities.
This factor therefore relates to positive affectivity. High scorers are likely to
be sociable, dominant and attention-seeking. Low scorers will tend to be
relatively unfriendly, detached and submissive.
Factor 1 - Impulsivity (1) (33 items). This scale measures the first higher
order factor of the APQ, which is virtually identical to the first SHAPS fac-
tor of impulsivity-aggression. The scale consists of items from the SC, PA,
RE, AG, and DE scales, and it is therefore a general factor of awareness and
expression of hostile, rebellious, and aggressive impulses. High scores
reflect impulsivity, antagonism, and noncompliance, while low scores relate
to control and conformity. The scale has 24 items in common with the ear-
lier B factor scale (Blackburn, 1987). Since accumulated data suggest that
high scores are more likely to be associated with overt “belligerent” beha-
viour when scores on the second factor (W) are low, the label “impulsivity”
is preferred.
Factor 2 - Withdrawal (W) (24 items). This scale measures the second
APQ factor which corresponds to the SHAPS withdrawal-sociability factor.
The scale is very similar to the earlier W factor scale, with which it shares
21 items. These are drawn from the AV and EX scales. Higher scores reflect
social withdrawal and anxiety, and submissiveness, while lower scores are
associated with sociability, self-confidence, and assertiveness.
Reliability and Validity of the APQ
The validity of the APQ scales as measures of deviant personality traits
was examined in terms of their differentiation between offenders and nor-
mals, and within the patient sample, between mentally ill and nonmentally
ill (legally psychopathic disorder) patients, Following the recommendationsTHE ANTISOCIAL PERSONALITY QUESTIONNAIRE 99
of Briggs and Cheek (1986) that new scales should be shown to be valid at
the conceptual, self-report, and behavioural levels, validity was also deter-
mined from the relationship of APQ scales to measures of personality disor-
der, observer ratings of social behaviour, and criminal history.
Reliability
Scale reliabilities (Cronbach's alpha) range from .77 to .87 for the total
sample (median = .86), from .79 to .88 for patients (median = .83), and from
-75 to .84 for normals (median = .80), indicating satisfactory internal con-
sistencies for all scales. Average inter-item correlations within scales range
from .14 to .28 for the total sample, from .15 to .28 for patients, and from
.13 to .26 for normals, indicating acceptable homogeneity for all scales.
Differentiation of Offenders and Normals
Table 3 shows the means and SDs on APQ scales of the mentally disor-
dered offender and normal samples. Also shown are data from a small group
of offenders (N = 29) referred for assessment by a probation service. This
group is younger than the other two samples (mean age 26.55; SD = 10.41),
and its representativeness of offenders on probation is unknown. However, it
provides some indication of the scores of noninstitutionalised offenders.
Table 3
Scores of Mentally Disordered Offenders, Normals and Probationers on APQ Scales
Patients Normals, Probationers
(N=499) (N=238) (W=29)
Scales M (SD) M (SD) M (SD)
Basic scales
Self-Control (SC) 10.44 (4.46) 7.66 (3.69) 7.03 (2.89)
Self-Esteem (SE) 7.92 (4.99) 483 (3.79) 941 (4.38)
Avoidance (AV) 7.10 (4.35) 455° (385) 748 (4.09)
Paranoid Suspicion (PA) 571 (4.42) 3.29 (3.09) 938 (3.71)
Resentment (RE) 9.44 (4.69) 6.65 (4.06) 1255 (291)
Ageression (AG) 9.13 (4.30) 933 (4.02) 1293 (4.02)
Deviance (DE) 1021 (4.27) 7:38 (3.66) 1389 (3.04)
Extraversion (EX) 1154 (4.28) 13.20 (3.81) 1121 (3.43)
Factor scales
Impulsitivy (1) 15.71 (6.96) 1355 (5.86) 22.10 (5.23)
Withdrawal (W) 9.88 (5.71) 71S (491 10.69 (5.04)
Nore. APQ = Antisocial Personality Questionnaire100 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
The APQ scales differentiate the mentally disordered offender and normal
groups at a highly significant level, and there are differences on all scales
except AG. The patient group is significantly more controlled, avoidant,
paranoid, resentful, socially deviant, and introverted than the normal
sample, and has lower self-esteem. Patients also score higher than normals
on both the I and W factor scales. The probationers also differ significantly
from normals in being more introverted and more deviant on all other scales
except SC. Although patients do not differ from normals on the aggression
(AG) scale, both psychopathic disorder patients and probationers score
higher on this scale than normals or mentally ill offenders
Within the patient sample, psychopathic disorder patients are also less
self-controlled, and more deviant, resentful and impulsive than the mentally
ill. There are no significant differences on the remaining five scales, consis-
tent with findings of substantial overlap between the two groups in persona-
lity dysfunction (Blackburn, Crellin, Morgan, & Tulloch, 1990). However,
the scores of the psychopathic disorder patients are very similar to those of
the probationer sample, and the only significant difference between these
two groups is a higher score of the probationers on paranoid suspicion (p <
.05). The implication that low self-control, and high paranoid suspicion,
resentment, aggression, deviance and impulsivity are common among non-
mentally ill offenders is congruent with evidence in the literature on the
salient characterisitics associated with criminality (e.g., Caspi et al., 1994).
The APQ and Personality Disorder
Further evidence for validity was obtained from a subgroup of 130 of the
mentally disordered offenders who had completed self-report measures of
personality disorder (MCMI-1). Correlations with MCMI-I scales generally
support the validity of all APQ scales as measures of personality deviation.
Highest correlations of APQ scales with MCMI-1 scales are: Self-Control
with Compulsive (.62); Self-Esteem with Avoidant (.77) and Passive-
Aggressive (.64); Avoidance with Avoidant (.79) and Schizoid (.74);
Paranoid Suspicion with Passive-Aggressive (.71), Avoidant (.69), and
Compulsive (-.64): Resentment with Passive-Aggressive (.65) and
Compulsive (-.60); Aggression with Passive-Aggressive (57), Compulsive
(-.61), and Antisocial (.41); Deviance with Passive-Aggressive (.71) and
Compulsive (-.70); and Extraversion with Histrionic (.73) Narcissistic (.71),
and Schizoid (-.67).
The prominence of correlations with the MCMI-1 Avoidant and Passive-
Aggressive scales is consistent with findings that these two scales have high
loadings on the two largest factors of the MCMI-1, which in turn correspondTHE ANTISOCIAL PERSONALITY QUESTIONNAIRE 101
broadly to the two factors in the SHAPS and APQ. This is seen in the corre-
lations of the two APQ factor scales, Impulsivity correlating positively with
Passive-Aggressive (.70) and Paranoid (.48) and negatively with Compulsive
(-.70), while Withdrawal correlates positively with MCMI-1 Avoidant (.74),
Schizoid (.74), and Schizotypal (.70) and negatively with Histrionic (-.72)
and Narcissistic (-.68). There is clearly substantial overlap between the two
sets of measures, all personality disorder scores on the MCMI-I being sig-
nificantly predictable from APQ scales. However, it should be emphasised
that these personality disorder categories are defined by Millon’s theory and
do not necessarily correspond to the personality disorder categories of DSM-
TI and DSM-IV.
APQ Scales and Observer Ratings
The Chart of Interpersonal Reactions in Closed Living Environments
(CIRCLE, Blackbur & Renwick, 1996) is an observer rating scale designed
to measure the Leary interpersonal circumplex (Leary, 1957; Wiggins,
1982). Eight scales measure the octants of the circle (Dominant, Coercive,
Hostile, Withdrawn, Submissive, Compliant, Friendly, Gregarious).
Correlations of APQ scales with these ratings of interpersonal style were
obtained for 100 patients.
All scales correlate significantly with at least one CIRCLE scale, and EX,
AV, AG and W attain correlations of between .4 and .5. Correlations are hig-
hest for APQ scales assumed to measure the most visible behavioural ten-
dencies (EX, W, AG, AV, DE), and lowest for scales related to beliefs (PA,
RE, SE). The pattern of correlations generally supports the validity of APQ
scales. Patients scoring high on EX, for example, are likely to be rated as
gregarious, dominant, and coercive, and as less submissive and withdrawn,
while the opposite pattern applies for those scoring high on AV and W. The
validity of AG is also well supported. High scorers are more likely to be
rated as coercive, dominant, gregarious, and hostile, and as less compliant
and submissive. High scores on DE and I are also significantly associated
with coercion, dominance, and noncompliant behaviour, consistent with the
content of the scales.
Because self-reports represent conceptions of self and observer ratings
are subject to perceptual biases, correlations between the two measurement
methods are invariably subject to attenuation. A significant degree of con-
vergence between self-report measures and behaviour ratings is therefore
good evidence for construct validity. The finding that four of the APQ
scales, including the withdrawal factor scale, have moderately strong corre-
lations with CIRCLE scales supports their validity as measures of social dis-
positions.102 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
Criminality and the APQ
Criminal records were available for 132 patients in the validation sample.
Data were extracted on age at first conviction, total number of convictions,
and total convictions for property acquisition (burglary, theft), violence
(excluding rape), robbery, sex offences (including rape), criminal damage,
and arson. A total assault measure was also obtained by combining convic-
tions for violence, robbery and sexual assault.
Age at first conviction is one of the strongest predictors of persistent cri-
minality (Tracy, Wolfgang, & Figlio, 1986), and it correlates significantly
(p < .05 or greater) with DE (-.44), I (-.34), AG (-.31), PA (-.23), and RE
(-.22). Higher scores on these scales are therefore associated with earlier cri-
minal careers. Higher scores on SC, in contrast, are associated with the later
appearance of criminal behaviour, as indicated by a positive correlation of
30. Total number of convictions correlates with DE (.32), I (21), and RE
(.20), and these scales also correlate significantly with number of acquisiti-
ve offences (.33, .20, and .18, respectively). DE also correlates with number
of sex offences (.18), while AG correlates with convictions for violence
(17), robbery (.17), and total assaults (.21). Criminal damage and arson
were not associated with any APQ scale.
Overall, DE has the strongest correlations with general criminality, sup-
porting its validity as a measure of socially deviant behaviour, while RE and
I show similar though less pronounced patterns. Although small, the signifi-
cant correlations of AG with violence, robbery, and total assaults provide
further evidence for the validity of this scale. The SE, AV, EX, and W scales
are not directly related to any aspect of criminality, although these traits may
modulate the expression of antisocial traits.
Factor Dimensions and Profile Types
Principal components analyses of the eight basic APQ scales yield simi-
lar results in the mentally disordered offender and normal samples, two sig-
nificant factors accounting for 74% of the variance in the patient data, and
70% for normals (Table 4). The first factor contrasts aggression (AG),
resentment (RE), deviance (DE) and paranoid beliefs (PA) with self control
(SC), and represents a dimension of impulsivity and hostile beliefs versus
control and conformity. The second factor contrasts avoidance (AV) and.
poor self esteem (SE) with extraversion (EX). The correspondence of these
factors with the impulsive aggression and withdrawal factors of the SHAPS
is confirmed by correlations of APQ factor scores with factor scores derived
from the SHAPS. Within both patient and normal samples, these are .95 and
.94, respectively.THE ANTISOCIAL PERSONALITY QUESTIONNAIRE. 103
Table 4
APQ Rotated Factors
Patients (N=499) Normals (N=238)
Factors Factors
APQ scale 1 u 1 u
sc 19 old -84 07
SE 37 66 49 66
AV 35 7 19 88
PA 3 39 nN 44
RE 84 08 Is 25
AG 87 -.04 84 16
DE 82 18 78 24
EX 18 -.88 19 85
% variance 47 27 42 28
The I and W scales were constructed to provide measures of these two
APQ factors. Correlations of I and W with scores on the two factors are .95
and .95 within the patient group and .92 and .94 within the normal group.
However, I and W are not entirely independent of each other, as shown by
correlations of .22 (p < .001) in patients and .16 (p < .02) in normals.
Nevertheless, for practical purposes, I and W provide good measures of the
two factors and summarise most of the variance in the APQ scales.
APQ Profile Paiterns
Cluster analyses (Ward's method) of APQ factor scores and scores on the
land W factor scales of the mentally disordered offender sample reproduce
the four profile patterns found previously in analyses of the SHAPS, Table 5
shows the mean profiles resulting from analysis of I and W scores. Primary
and secondary psychopaths score high on I, but differ on W. Controlled and
inhibited groups score low on I, but again differ on W. The four profiles em-
Phasise different orientations towards self and others and different styles of
coping with the social environment.
In terms of modal or prototypic traits, primary psychopaths are extraver-
ted and dominant (EX, AV), aggressive (AG) moderately self-controlled
(SC), mistrusting (RE), have a history of deviant behaviour and conflict in
relationships (DE), and are relatively untroubled by anxiety or self-criticism
(SE). Secondary psychopaths are also aggressive, low in self-control, mis-
trusting, and socially deviant, but are socially anxious and submissive (EX,
AY), moody and lacking in self-esteem (SE), and are extremely suspicious
and vigilant for threat (PA). Controlled personalities are overcontrolled and104 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
Table 5
Means of APQ Profile Patterns
Primary Secondary Controlled Inhibited
psychopaths _psychopaths
Scale (N=219) (N=58) (W=120) (N=102)
Fe eee
Basic scales
sc 9.29 sal 14,32 1119
SE 7.94 14.13 2.16 10.31
AV 6.17 13.16 273 10.80
PA 6.49 1205 178 5.07
RE 11.23 1467 463 7.20
AG 11.08 13.34 5.28 7.09
DE 11.79 15.05 593 9.10
EX 13.93 674 13.48 7.19
Factor scales
1 19.47 24.14 795 11.95
w 1.55 18.36 491 15.88
—_—
conforming (SC, AG, DE), sociable and outgoing (EX, AV), self-accepting
(SE), and deny interpersonal mistrust or suspicion (RE, PA). Inhibited per-
sonalities are also overcontrolled, compliant and relatively trusting (SC, AG.
RE), although reporting somewhat more deviant histories (DE) and self-cri-
ticism (SE), and are introverted, timid, and withdrawn (EX, AV). Most
patients in the inhibited group have been identified as suffering from serious
mental disorder.
Criminal history data of the four groups within the validation sample are
shown in Table 6. Planned comparisons indicate that primary and secondary
psychopaths (combined) differ from the controlled and inhibited groups
(combined) in being younger at first conviction (p < .0001), and in having
more convictions overall (p < .01), more custodial sentences (p < .01), and
more convictions for burglary and theft (p < .05). They also tend to have
more convictions for assaultive offences (p < .06). Secondary psychopaths
have had more custodial sentences than primary psychopaths (p < .05), but
otherwise have comparable criminal histories. There are no differences
between controlled and inhibited groups. These data confirm earlier findings
that primary and secondary psychopaths have longer and more extensive cri-
minal careers (Blackburn, 1975a), and represent two forms of antisocial per-
sonality. Contrary to previous findings, however, primary psychopaths are
similar to secondary psychopaths in their history of violent crimes.THE ANTISOCIAL PERSONALITY QUESTIONNAIRE 105
Table 6
Mean Criminal Convictions of APQ Groups
$s
Primary Secondary Controlled Inhibited
psychopaths psychopaths
Convictions (N51) (N=24) (N=30) (N=27)
Age at first 17.20 16.50 2207 22.33
Total 13.57 16.46 9.00 796
Custodial sentences 2.39 4.08 1.90 1.63
Burglary/theft 6.10 9.08 417 4.26
Robbery 0.24 0.17 out 0.07
Viotence 231 2.29 197 141
Sex offences o7 1,00 0.47 0.22
Criminal damage 0.84 0.83 087 0.83
Arson 0.37 0.08 0.10 037
Total assaults 3.08 3.04 2.60 1.70
Personality Questionnaires in the Assessment of Offenders
Although many psychologists continue to be sceptical about the utility of
personality trait description, there is now ample evidence that personality is
a risk factor for both antisocial behaviour (Andrews & Wormith, 1989;
Blackbur, 1993) and psychopathology (Krueger et al., 1996). In the light of
recent developments in personality research, it has also been suggested that
personality trait assessment should be a required component of any clinical
evaluation (Harkness & Lillienfeld, 1997). However, several practical and
theoretical issues need to be considered in assessing personality in offenders.
The first is the context of assessment. Offenders are not voluntary clients
and forensic assessment is usually undertaken for the benefit of society
rather than the individual offender. Violent and sexual offenders in particular
frequently deny their offence or minimise its consequences, and are reluc-
tant to become engaged in assessment or clinical intervention. Effective
assessment therefore requires the development of trust between psychologist
and offender. The interpretation of assessment data also needs to take
account of the pressures on an offender to avoid disclosure of deviant attri-
butes. For example, assessment of suitability for parole or release from
detention is more likely to elicit defensive denial or impression management
than is assessment undertaken for treatment planning or research purposes.
A second issue is the choice of method of personality assessment.
Psychologists have traditionally relied most commonly on self-report
questionnaires, but recent interest in the identification of personality disor-
ders has led to the development of a number of structured interviews for106 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
assessing personality deviation (Zimmerman, 1994). Assessment of psycho-
pathy with Hare’s PCL-R also takes this form. Because structured interviews
are more flexible and permit more in-depth probing of an individual's histo-
ry, it is sometimes assumed that they provide more valid assessments of per-
sonality than questionnaires. However, interviews are subject to problems of
terviewer bias and halo effects, and in relying on self-reports, they share
methodological problems with questionnaires arising from an interviewee's
defensiveness or poor self awareness.
It is also sometimes argued that offenders cannot be trusted to answer
self-reports honestly (e.g., Hare, 1991). However, there is little evidence that
this is invariably a problem or that interviews successfully detect dissimula-
tion, The assumption that people have a “true” personality that can be detec-
ted more readily by observers than by the self is also questionable on both
methodological and philosophical grounds. Personality assessment through
self-report is concerned less with eliciting verifiable facts than with beliefs
about the sort of person one is, and is dependent on a person’s self-concept
and self-presentation (Costa & McCrae, 1992; Hogan & Nicholson, 1988).
Given that the demands of personality questionnaires and interviews are
similar in this respect, there seems no reason why offenders should present
themselves differently in response to one rather than the other. Nevertheless,
except for research purposes, personality questionnaires should not be relied
on as the sole source of assessment. They should normally form part of a
multimethod battery of measures which includes interview data and obser-
vations by others, such as rating scales completed by staff (e.g., Blackburn
& Renwick, 1996).
One further issue is the inherent limitation of trait assessment in explai-
ning and predicting behaviour. Traits are summary descriptions of tenden-
cies or dispositions which are likely to be manifest in average or typical
behaviours across time and relevant settings. Knowledge of a person's dis-
positions provides clues as to person variables (e.g., generalised expecta-
tions, attitudes towards self and others, coping styles or strategies) that
mediate reactions to particular events or which generate problematic inter-
actions. However, they permit only probabilistic statements about future
behaviour, and whether dispositions are realised in specific acts depends on
the situation and the person’s current state. In predicting violence, for exam-
ple, we may at best judge the strength of an aggressive disposition, and can
rarely forecast the situational contexts in which that disposition may eventu-
ate in an extreme act of aggression (Blackburn, 1993). The identification of
an individual's predominant traits is therefore simply a first step in under-
standing their behaviour.THE ANTISOCIAL PERSONALITY QUESTIONNAIRE. 107
Conclusions
Although the APQ is a relatively new instrument, its scales represent psy-
chometrically refined versions of the earlier SHAPS scales, and the APQ
retains the dimensional structure of the SHAPS. Findings from research with
the SHAPS should therefore be generalisable to the APQ. For example, the
correlations of the APQ I and W factor scales with the main factors of per-
sonality disorder in the MCMI-1 (.75 and .81) are virtually identical to those
of the SHAPS B and W factor scales (.77 and .80), despite a number of chan-
ges in item content. It is reasonable to expect that the significant correlation
of the SHAPS B scale with the PCL-R will similarly be reproduced by the
APQ I scale. At the same time, the greater homogeneity of the APQ scales
is likely to improve validity. For example, the correlation of the APQ
Aggression scale with observer ratings of coercive interpersonal style is sub-
stantially higher than that of the SHAPS equivalent (.41, p < .0001, vs .21,
p< .05,N = 100).
Data obtained to date indicate that the APQ scales provide a valid basis
for inferring significant intrapersonal and interpersonal characteristics
among offenders. They may therefore aid the formulation of an individual’s
problems, identify targets of intervention, such as social anxiety, aggression
or hostile attributions, provide a baseline against which to evaluate change,
and contribute to the prediction of future behaviour. The four profile patterns
identified previously in the SHAPS, and reproduced by the APQ, are also
relevant to differential classification (Andrews, Bonta, & Hoge, 1990).
While further research is needed to establish the generalisability of the APQ
to different offender populations, such as women and adolescents, the utili-
ty of the questionnaire for forensic assessment scems promising.
References
American Psychiatric Association. (1987). Diagnostic and statistical manual of men-
tal disorders (3rd ed., rev.). Washington, DC: Author.
Andrews, D. A., & Wormith, J. S. (1989). Personality and crime: Knowledge destruc-
tion and construction in criminology. Justice Quarterly, 6, 289-309.
Andrews, D. A., Bonta, J., & Hoge, R. D. (1990). Classification for effective rehabi-
litation: Rediscovering psychology. Criminal Justice and Behavior, 17, 19-52.
Beckett, R., Beech, A., Fisher, D., & Fordham, A. S. (1994). Community-based tre-
aiment for sex offenders: An evaluation of seven treatment programmes.
London: Home Office
Black, D. A., & Spinks, P. (1985). Predicting outcomes of mentally disordered and
dangerous offenders. In D. P. Farrington & R. Tarling (Eds.), Prediction in cri-
minology (pp. 48-69). Albany: State University of New York Press,108 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE.
Blackburn, R. (1971a). MMPI dimensions of sociability and impulse control
Journal of Consulting and Clinical Psychology, 37, 166.
Blackburn, R. (1971b). Personality types among abnormal homicides. British
Journal of Criminology, 11, 14-31
Blackburn, R. (1972). Dimensions of hostility and aggression in abnormal offenders.
Joumal of Consulting and Clinical Psychology, 38, 20-26.
Blackburn, R. (1974). The development and validation of scales to measure hostility
and aggression. Special Hospitals Research Reports, No. 12. London: Special
Hospitals Research Unit.
Blackburn, R. (1975a). An empirical classification of psychopathic personality.
British Journal of Psychiatry, 127, 456-460.
Blackburn, R. (1975b). Aggression and the EEG: A quantitative analysis. Journal of
Abnormal Psychology, 84, 358-365.
Blackburn, R. (1978). Psychopathy, arousal, and the need for stimulation. In R. D.
Hare & D. Schalling (Eds.), Psychopathic behavior: Approaches to research
(pp. 157-164). Chichester: Wiley.
Blackburn, R. (1979a). Psychopathy and personality: The dimensionality of self-
report and behaviour rating data in abnormal offenders. British Journal of
Social and Clinical Psychology, 18, 111-119.
Blackburn, R. (1979b). Cortical and autonomic arousal in primary and secondary
psychopaths. Psychophysiology, 16, 153-160.
Blackburn, R. (1982). The Special Hospitals Assessment of Personality and
Socialisation (SHAPS). Unpublished manuscript, Park Lane Hospital,
Liverpool, UK.
Blackburn, R. (1986). Patterns of personality deviation among violent offenders:
Replication and extension of an empirical taxonomy. British Journal of
Criminology, 26, 254-269.
Blackburn, R. (1987). Two scales for the assessment of personality disorder in anti-
social populations. Personality and Individual Differences, 8, 81-93.
Blackburn, R. (1993). The peewee’ of criminal conduct: Theory, research and
practice. Chichester: Wiley.
Blackburn, R. (1996). Replicated personality disorder clusters among mentally dis-
ordered offenders and their relation to dimensions of personality. Journal of
Personality Disorders, 10, 68-81.
Blackburn, R. (1998). Psychopathy and personality disorder: Implications of interper-
sonal theory. In D. J. Cooke, A. E. Forth, & R. D. Hare (Eds.), Psychopathy:
Theory, research and implications for society (pp. 269-301). Amsterdam: Kluwer.
Blackburn, R., & Coid, J. W. (1998). Psychopathy and the dimensions of personali-
ty disorder in violent offenders. Personality and Individual Differences, 25,
129-145.
Blackburn, R., & Coid, J. W. (1999). Empirical clusters of DSM-III personality dis-
orders in violent offenders. Journal of Personality Disorders, 13, 18-34.
Blackbur, R., & Fawcett, D. J. (1996). Manual for the Antisocial Personality
Questionnaire. Unpublished manuscript, University of Liverpool, UK.
Blackburn, R., & Fawcett, D. J. (1999). The Antisocial Personality Questionnaire: An
inventory for assessing deviant traits in offender populations. European Journal
of Psychological Assessment 15, 14-24.THE ANTISOCIAL PERSONALITY QUESTIONNAIRE 109
Blackburn, R., & Lee-Evans, J. M. (1985). Reactions of primary and secondary psy-
chopaths to anger-evoking situations. British Journal of Clinical Psychology,
24, 254-269.
Blackburn, R., & Maybury, C. (1985). Identifying the psychopath; The relation of
Cleckley’s criteria to the interpersonal domain. Personality and Individual
Differences, 6, 375-386.
Blackburn, R., & Renwick, S. J. (1996). Rating scales for measuring the interpersonal
circle in forensic psychiatric patients. Psychological Assessment, 8, 76-84.
Blackburn, R., Crellin, M. C., Morgan, E. M., & Tulloch, R. M. B, (1990).
Prevalence of personality disorders in a special hospital population, Journal of
Forensic Psychiatry, 1, 43-52.
Briggs, S, R., & Check, J. M. (1986). The role of factor analysis in the development
and evaluation of personality scales. Journal of Personality, 54, 166-208,
Buss, A. H. (1961). The psychology of aggression. New York: Wiley.
Caspi, A., Moffitt, T. E., Silva, P. A., Stouthamer-Loeber, M., Krueger, R. F., &
Schmutte, P. S. (1994). Are some people crime-prone? Replications of the per-
sonality-crime relationship across countries, genders, races, and methods.
Criminology, 32, 163-195.
Clark, D. (1996). Personal communication.
Clark, L. A., Livesley, W. J., Schroeder, M.. L., & Irish, S. L. (1996). Convergence
of two systems for assessing specific traits of personality disorder.
Psychological Assessment, 8, 294-303.
Cleckley, H. (1976). The mask of sanity (6th ed.). St. Louis: Mosby.
Como, P. G. (1977). The use of multiple personality variables to predict recidivism.
Unpublished MA thesis, University of Texas.
Corah, N. (1964). Neuroticism and extraversion in the MMPI: Empirical validation
and exploration, British Journal of Social and Clinical Psychology, 3, 168-174.
Costa, P. T., & McCrae, R. R. (1992). Normal personality assessment in clinical prac-
tice: The NEO personality inventory. Psychological Assessment, 4, 5-13.
Davis, H. M. (1974). Psychometric prediction of institutional adjustment: A valida-
tion study. British Journal of Social and Clinical Psychology, 13, 269-276.
Dolan, M. C. (1996). Serotonergic function in personality disordered offenders. In D.
J. Cooke, A. E. Forth, J. Newman, & R. D. Hare (Eds.), International perspec-
tives on psychopathy: Issues in criminological and legal psychology (No. 24,
pp. 30-34). Leicester: British Psychological Society.
Foulds, G. A. (1971). Personality deviance and personal symptomatology.
Psychological Medicine, 1, 222-233.
Giedt, F. H., & Downing, L, (1961). An extraversion scale for the MMPI. Journal of
Clinical Psychology, 17, 156-159.
Gillham, R. (1978). An investigation of imagery in psychopathic delinquents
Unpublished BSc dissertation, University of Aberdeen, Scotland.
Gray, J. A. (1987). Perspectives on anxiety and impulsivity: A commentary. Journal
of Research in Personality, 21, 493-509,
Hare, R. D. (1991). The Hare Psychopathy Checklist-Revised. Toronto: Multi-Health
Systems.
Hare, R. D. (1996). Psychopathy: A construct whose time has come. Criminal Justice
and Behavior, 23, 25-54,110 THE ANTISOCIAL PERSONALITY QUESTIONNAIRE
Harkness, A. R., & Lillienfeld, S. O. (1997). Individual differences science for treat-
ment planning: Personality traits. Psychological Assessment, 9, 349-360.
Heilbrun, A. B., & Heilbrun, M. R. (1985). Psychopathy and dangerousness:
Comparison, integration, and extension of two psychopathic typologies. British
Journal of Clinical Psychology, 24, 181-95
Henderson, M. (1982). An empirical classification of convicted violent offenders.
British Journal of Criminology, 22, 1-20.
Hogan, R., & Nicholson, R. A. (1988). The meaning of personality test scores.
American Psychologist, 43, 621-626.
Howard, R. C. (1984). The clinical EEG and personality in mentally abnormal offen-
ders. Psychological Medicine, 14, 569-580.
Howard, R. C., Fenton, G. W., & Fenwick, P, B.C, (1984). The contingent negative
variation, personality, and antisocial behaviour. British Journal of Psychiatry,
144, 463-474,
Kassebaum, G. C., Couch, A. S., & Slater, P. E. (1959). The factorial dimensions of
the MMPI. Journal of Consulting Psychology, 23, 226-236
Kosson, D. S., Smith, S. S., & Newman, J. P. (1990). Evaluating the construct validi-
ty of psychopathy on black and white male inmates: Three preliminary studies.
Journal of Abnormal Psychology, 99, 250-259.
Kroner, D. (1997). Personal communica
Kreger, R. F, Caspi, A., Moffitt, T. E., Silva, P. A., & McGee, R. (1996). Personality
traits are differentially linked to mental disorders: A mulitrait-multidiagnosis study
of an adolescent birth cohort. Journal of Abnormal Psychology, 105, 299-312.
Kuriychuk, M. (1990). The assessment of psychopathy and risk-taking behavior.
Unpublished doctoral dissertation, Queen’s University, Kingston, Ontario.
Leary, T. (1957). Interpersonal diagnosis of personality. New York: Ronald Press.
Lykken, D. T. (1957). A study of anxiety in the sociopathic personality. Journal of
Abnormal and Social Psychology, 35, 6-10.
McCrae, R. R., & Costa, P. T, (1986). Clinical assessment can benefit from recent
advances in personality psychology. American Psychologist, 41, 1001-1003.
McGurk, B. J. (1978). Personality types among normal homicides. British Journal of
Criminology, 18, 146-161.
McGurk, B. J., & McGurk, R. E. (1979). Personality types among prisoners and pris
on officers. British Journal of Criminology, 19, 31-49.
Megargee, F. I. (1966). Undercontrolled and overcontrolled personality types in
extreme antisocial aggression. Psychological Monographs, 80, Whole No. 611.
Millon, T. (1983). Millon Clinical Multiaxial Inventory (3rd. ed.). Minneapolis:
Interpretive Scoring Systems.
Needs, A. (1986). Personal communication.
O'Kane, A., Fawcett, D., & Blackburn, R. (1996). Psychopathy and moral reasoning:
Comparison of two classifications. Personality and Individual Differences, 20,
505-514.
Peterson, D. R., Quay, H. C., & Cameron, G. R. (1959). Personality and background
factors in juvenile delinquency as inferred from questionnaire responses.
Journal of Consulting Psychology, 23, 295-299.
Pollock, P. H. (1999). When the killer suffers: Post-traumatic stress reactions follo-
wing homicide. Legal and Criminological Psychology, in press.THE ANTISOCIAL PERSONALITY QUESTIONNAIRE 1
Porter, S. (1996). Without conscience or without active conscience? The etiology of
psychopathy revisited. Aggression and Violent Behavior, I, 179-189.
Renwick, S. J., & Emler, N. (1991). The relationship between social skills deficits
and juvenile delinquency. British Journal of Clinical Psychology, 30, 61-12
Salekin, R. T., Rogers, R., & Sewell, K. W. (1996). A review and meta-analysis of the
Psychopathy Checklist and the Psychopathy Checklist-Revised: Predictive validi-
ty of dangerousness. Clinical Psychology: Science and Practice, 3, 203-215.
Stein, K. B. (1968). The TSC scales: The outcome of a cluster analysis of the 550
MMPI items. In P. McReynolds (Ed.), Advances in psychological assessment
(Vol. 1, pp. 80-104). Palo Alto: Science and Behavior Books.
Tracy, P., Wolfgang, M., & Figlio, R. (1986). Delinquency in a birth cohort 2: A com-
parison of the 1945 and 1958 Philadelphia birth cohorts. Washington, DC:
National Institute of Juvenile Justice and Delinquency.
Watson, D., & Clark, L. A . (1984), Negative affectivity: The disposition to expe-
rience negative affective states. Psychological Bulletin, 96, 465-490,
Welsh, G. S. (1956). Factor dimensions A and R. In G. S. Welsh & W. G. Dahlstrom
(Eds.), Basic readings on the MMPI in psychology and medicine (pp. 106-129).
Minneapolis: University of Minnesota Press.
Widom, C. S. (1978). An empirical classification of the female offender. Criminal
Justice and Behavior, 5, 35-52
Wiggins, J. S. (1982). Circumplex models of interpersonal behaviour in clinical psy-
chology. In P. C. Kendall & J. N. Butcher (Eds.), Handbook of research
methods in clinical psychology (pp. 183-221). New York: Wiley.
Willner, A. H., & Blackburn, R. (1988). Interpersonal style and personality deviation
British Journal of Clinical Psychology, 27, 273-274.
Wilson, J. Q., & Herrnstein, R. S. (1985). Crime and human nature. New York:
Simon & Schuster.
Zimmerman, M. (1994). Diagnosing personality disorders: A review of issues and
research methods. Archives of General Psychiatry, 51, 225-245.
Received April, 1999