Measuring cognitive schemas in sexual offenders with self-report and implicit associations

Offense-supported beliefs are common in sex offenders, contribute to the onset
of sexual offending and affect the risk of re-offending (Ó Ciardha & Ward,
2013; Hanson & Morton-Bourgon, 2004; Helmus, Hanson, Babchishin, & Mann, 2013).
According to the literature, sexual offenders* offense-supported beliefs are
the product of so-called Implicit Theories (ITs) that sexual offenders develop
and use to make sense of their social world, guide their choices and pursuit
their goals (Ward & Keenan, 1999; Polaschek & Ward, 2002; Ward, 2000). There
are two types of ITs for sexual offenders (1) general antisocial oriented
theories regarding the nature of the world and people (i.e., Entitlement,
Uncontrollability, Dangerous World) and more general middle-level theories
related to specific people and type of victim (i.e., Children as sexual beings,
Nature of harm, Women are unknowable, Women as sexual objects). Traditionally,
ITs are measured by indirect measurements (self-report/clinical interviews),
which can give rise to methodological concerns and questions. The major
disadvantages are the focus on almost only middle-level beliefs (Gannon, Keown,
& Rose, 2008), possible response bias (e.g., Gannon, Keown, & Polaschek, 2007;
Mann, Webster, Wankeling, & Marshall, 2007) and the possible inability to
measures unconscious processes (e.g., Fazio & Olson, 2003; Gawronski &
Bodenhausen, 2007; Ward & Keenan, 1999). To overcome some of these problems,
researchers have developed latency based indirect measures (i.e., measures to
assess constructs under automaticity) in which participants are asked to
respond as quickly as possible to stimuli that appear on a computer screen. The
basic assumption of these measures is that one should be faster and make fewer
errors if the condition (i.e., presented stimuli) corresponds with their
beliefs. Some indirect measures do this by assessing strengths of cognitive
associations by comparing reaction times after offering associative stimuli to
different pairings of concepts (i.e., Implicit Association Task). Other
measures require the participant to answer in accordance with specific beliefs
(i.e., Implicit Relational Assessment Procedure (IRAP); Relational Response
Task (RRT)). Because the use of indirect measures in the field of forensics is
rather new (i.e., IRAP, RRT), it is important to investigate to what extent
indirect measures are capable of measuring ITs in terms of convergent,
discriminant * (different offenders, social desirability), predictive and
incremental validity (i.e., over and above direct measures) in forensic
patients.

The study aims to fill gaps in current research regarding the measurements of
ITs by (1) developing a self-report questionnaire covering all ITs, (2)
investigate to what extent indirect measures are capable of capturing ITs in
terms of test-retest, discriminative, predictive (Sexual interest, recidivism
risk) and incremental validity (over and above self-report measures). Secondary
objectives: (1) look at differences on the QUITSO and Indirect measures between
type of sexual offender, (2) investigate the role of social desirability and
age as covariates.

Primary objectives: (1) develop valid measurements using multiple methods
(self-report (i.e., QUITSO), Indirect measures) to identify ITs in child
molesters, pedophiles and rapists and to examine the differences between child
molesters, pedophiles, rapists and violent offenders (control group) on these
measures.
Secondary objectives: (1) look at differences on the QUITSO and IAT between
type of sexual offender, (2) investigate the role of social desirability and
age as covariates.

A quasi-random experimental design is used in which several measures of IT
(i.e., QUITSO and indirect measures), of different types of sexual offenders
are developed, examined and compared to existing measures (MOLEST scale,
RMAS). Based on both stratified randomization (type of sexual offender) and
treatment setting ( in- or outpatient), participants are asked to participate
in part A (QUITSO, MOLEST
scale, RMAS, BIDR-6) or part B (QUITSO, MOLEST scale, RMAS, BIDR-6, IAT, RRT,
IRAP, VT) of the study. Inpatient will be asked to participate in part B of the
study, outpatient in part A. Participants from the general population will be
asked to participate in Part A and B.

The current study will contribute to the further development of identifying
ITs, a risk factor for re-offending in sexual offenders, will attribute to the
development of the constructs used in current schema-based treatment methods
and will add possible research support for the use of these models.
Participation in the study will take an investment of approximately 60 minutes
(part A) or 1.5 hours (part B) from participants. The administration of
questionnaires and Indirect measures is a non-invasive, safe, painless
procedure not associated with any risks. While some emotional comfort can be
experienced, participants are informed about this on forehand and the
possibility to discuss these feelings are offered. Investigators do not expect
any serious negative consequences for participants who complete the
test-battery