Biased Symptom Reporting and Antisocial Behaviour in Forensic Samples: A Weak Link

Abstract

In two studies (one with 57 forensic inpatients and one with 45 prisoners) the connection between biased symptom reporting and antisocial behaviour is explored. The findings are as follows: 1) the association between symptom over-reporting and antisocial features is a) present in self-report measures, but not in behavioural measures, and b) stronger in the punitive setting than in the therapeutic setting; and 2) participants who over-report symptoms a) are prone to attribute blame for their offence to mental disorders, and b) tend to report heightened levels of antisocial features, but the reverse is not true. The data provide little support for the inclusion of antisocial behaviour (i.e. antisocial personality disorder) as a signal of symptom over-reporting (i.e. malingering) in the Diagnostic and Statistical Manual of Mental Disorders – Fifth Edition (DSM-5). The empirical literature on symptom over-reporting and antisocial/psychopathic behaviour is discussed and it is argued that the utility of antisocial behaviour as an indicator of biased symptom reporting is unacceptably low.

Key words:

• antisocial personality disorder
• malingering
• psychopathy
• response bias
• symptom validity

Acknowledgements

The authors wish to thank Jette Freeman for gathering a portion of the data.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. For scores of a certain index (e.g. the LSRPS, the SDAS-11, number of punitive actions) to have a diagnostic accuracy above chance level with regard to a dichotomous condition (e.g. symptom exaggeration, see note 3 below), the overlap between the index score distributions of the two conditions (e.g. honest and feigning) must be <50%. For example, if SDAS-11 scores are used as predictor of symptom exaggeration and the overlap between SDAS-11 scores of honest patients and feigning patients is 40%, then 60% of the patients can be classified correctly (i.e. 60% obtained scores that are unique to their group). Percentages of overlap between score distributions correspond to the magnitude of difference between the means of the score distributions (i.e. to Cohen's d; see Table 1 in Zakzanis, 2001). The sample size required for the conventional .80 statistical power to detect an effect that produces a diagnostic accuracy of at least 60% (i.e., d ≥ 1.2) is 54 when the base rate of symptom exaggeration is set at 10% (α = .05, one-tailed). The smallest groups in Studies 1 and 2 are the ones based on high SIMS scores (i.e. SIMS score >16; n = 5 and n = 6). The achieved power in comparisons of these groups with the groups that responded credibly to the SIMS is .81 for Study 1 and .85 for Study 2 (α = .05, one-tailed).

2. Studies were gathered from Niesten et al. (2015), and via a thorough search via Google Scholar with the search terms ‘psychopathy’ and ‘antisocial personality disorder’ combined with ‘malingering’, ‘feigning’, ‘simulation’, and ‘dissimulation’.

3. Symptom exaggeration is in fact not a dichotomous phenomenon; rather, it is dimensional, with feigned psychopathology stretched out along a continuum (Walters, Berry, Rogers, Payne, & Granacher, 2009; Walters, Rogers, et al., 2008). Nevertheless, practical decision-making often demands the assessment of feigned psychopathology to produce dichotomous outcomes (e.g. the honesty–feigning dichotomy).