ebook img

Biased Symptom Reporting And Antisocial Behavior In Forensic Samples PDF

20 Pages·2016·0.34 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Biased Symptom Reporting And Antisocial Behavior In Forensic Samples

Psychiatry, Psychology and Law ISSN: 1321-8719 (Print) 1934-1687 (Online) Journal homepage: http://www.tandfonline.com/loi/tppl20 Biased Symptom Reporting and Antisocial Behaviour in Forensic Samples: A Weak Link Alfons van Impelen, Harald Merckelbach, Isabella J. M. Niesten, Marko Jelicic, Benno Huhnt & Joost á Campo To cite this article: Alfons van Impelen, Harald Merckelbach, Isabella J. M. Niesten, Marko Jelicic, Benno Huhnt & Joost á Campo (2016): Biased Symptom Reporting and Antisocial Behaviour in Forensic Samples: A Weak Link, Psychiatry, Psychology and Law To link to this article: http://dx.doi.org/10.1080/13218719.2016.1256017 Published online: 12 Dec 2016. Submit your article to this journal View related articles View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=tppl20 Download by: [Maastricht University Library] Date: 15 December 2016, At: 01:55 Psychiatry,PsychologyandLaw,2016 http://dx.doi.org/10.1080/13218719.2016.1256017 BiasedSymptomReportingandAntisocialBehaviourinForensicSamples: AWeakLink AlfonsvanImpelena,HaraldMerckelbacha,IsabellaJ.M.Niestena,MarkoJelicica,Benno HuhntbandJoost(cid:1)aCampoc aDepartmentofClinicalPsychologicalScience,MaastrichtUniversity,TheNetherlands;bBerlin, Germany;cRadixForensicPsychiatricHospital,Heerlen,TheNetherlands In two studies (one with 57 forensic inpatients and one with 45 prisoners) the connection betweenbiasedsymptomreportingandantisocialbehaviourisexplored.Thefindingsareas 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 punitivesettingthaninthetherapeuticsetting;and2)participantswhoover-reportsymptoms 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 signalofsymptomover-reporting(i.e.malingering)intheDiagnosticandStatisticalManual of Mental Disorders – Fifth Edition (DSM-5). The empirical literature on symptom over- reportingandantisocial/psychopathicbehaviourisdiscussedanditisarguedthattheutility ofantisocialbehaviourasanindicatorofbiasedsymptomreportingisunacceptablylow. Keywords:antisocialpersonalitydisorder;malingering;psychopathy;responsebias;symp- tomvalidity. Introduction incentives’ (American Psychiatric Associa- Antisocial behaviour and deceptive tenden- tion,2013,p.726). Precisely because malingering is a form cies are considered to be core characteristics of deception – and because such behaviour of both antisocial personality disorder infringes social norms – the idea that malin- (ASPD; American Psychiatric Association, gering is strongly associated with ASPD and 1980, 2000, 2013) and psychopathy (Cleck- psychopathy has great intuitive appeal. ley, 1941, 1988; Hare, 1991, 2003; Hare, Forth, & Hart, 1989). A specific variant of Accordingly,theDSM–fromitsthirdedition deceptive behaviour is referred to in the onwards – assumes that antisocial behaviour Diagnostic and Statistical Manual of Mental is intimately linked to malingering. Indeed, theDSM-5liststhepresenceofASPDamong Disorders – Fifth Edition (DSM-5;American theindicationsthatwarrantheightenedsuspi- Psychiatric Association, 2013) as malinger- cion of malingering. However, in contrast to ing: ‘The intentional production of false or itsprimafacieplausibility,theempiricalsup- grossly exaggerated physical or psychologi- port for this idea is weak. In fact, early cal symptoms, motivated by external Correspondence:AlfonsvanImpelen,ForensicPsychologySection,DepartmentofClinicalPsychological Science, Maastricht University, PO Box 616, 6200 MD, Maastricht, The Netherlands. Tel. C31 43 388 1264.Email:[email protected] (cid:1)2016TheAustralianandNewZealandAssociationofPsychiatry,PsychologyandLaw 2 A.vanImpelenetal. reviews (Clark, 1997; DeMatteo & Edens, principally deceptive and manipulative in 2006) concluded that there is a paucity of nature. Hence, itstands toreason thatifanti- studies demonstrating the link between anti- social and psychopathic traits predispose to social behaviour and malingering. More symptom over-reporting (i.e. malingering), recently, Niesten, Nentjes, Merckelbach, and these features may also predispose to symp- Bernstein (2015) conducted a systematic tomunder-reporting. search by means of several databases and The extant literature about the relation confirmed the mixed findings in this domain: between symptom under-reporting and social of the seven studies found that explore desirability, on the one hand, and antisocial whether psychopathic and antisocial behav- and psychopathic traits, on the other, is even iour are related to symptom over-reporting, scarcer than that on symptom over-reporting fourfoundanassociation–albeitarelatively and antisocial or psychopathic traits. The weak one – (e.g. Heinze & Vess, 2005; investigation of Niesten et al. (2015) only Kucharski, Duncan, Egan, & Falkenbach, yielded two studies that are directly relevant: 2006), one did not find a relation (Pierson, Cima, van Bergen, and Kremer (2008), who Rosenfeld, Green, & Belfi, 2011), and two found no association, and Freeman and Sam- produced conflicting results (Cima & van son (2012), who found psychopathy to be Oorsouw, 2013; Sumanti, Boone, Savodnik, associated with less symptom under-reporting. & Gorsuch, 2006). Furthermore, a recent Additionally, the meta-analysis of Ray et al. meta-analysis of the relation between dis- (2013)showedthattheantisociallifestylefac- torted response styles and self-reported psy- tor (95% CI of weighted mean effect size: chopathic traits revealed a medium [¡.25, ¡.06]), but not the manipulative, cal- association between symptom over-reporting lous personality factor (95% CI: [¡.06, .05]) (i.e. malingering) and the antisocial lifestyle of psychopathy is negatively related to symp- factor(95%CIofweightedmeaneffect size: tomunder-reportingandsocialdesirability. [.23, .40]), but not the manipulative, callous One explanation for the conflicting find- personalityfactor(95%CI:[.00,.14])ofpsy- ings in this research domain is that the links chopathy(Rayetal.,2013). between biased symptom reporting and anti- Some authors (e.g. MacNeil & Holden, social features are context dependent. Thus, 2006) have speculated that high levels of psy- prison inmates may feign psychiatric symp- chopathy are associated with greater profi- toms in an attempt to be transferred from ciency in successful (i.e. undetected) faking. prison to the relatively mild conditions of a Even if psychopathic traits would confer no forensic psychiatric hospital. Once in a psy- aptitude for malingering directly, they could chiatric hospital, they may exaggerate their still lead to gains in proficiency through prac- mentalfitnesstoreducemandatorytreatment. tice,astheymaypromptindividualstoengage Similarly,defendantsmayfeignsymptomsin in malingering more frequently. However, an attempt to reduce their criminal responsi- there is hardly any support for the hypothesis bility, yet employ symptom under-reporting thatantisocialandpsychopathictraitsfosterthe and social desirability post-conviction to abilitytomalinger(e.g.Marionetal.,2013;for obtain privileges, probation, or parole. Like- anoverview,seeNiestenetal.,2015). wise,plaintiffsmayfeignparticularsymptom Distortedsymptomreportsduringclinical constellations (e.g. post-traumatic stress, assessmentarenotlimitedtotheexaggeration burnout, chronic pains) in the service of a ofsymptoms;they may alsotake theform of compensation claim, while simultaneously the denial of such symptoms, as well as the denying genuine problems (e.g. substance exaggeration of positive qualities or indica- use, impulsivity, compulsivity) to make a tors of good health (i.e. social desirability). favourable impression on judicial decision- Much like malingering, symptom under- makers (e.g. Cima & van Oorsouw 2013; reporting and social desirability are Niestenetal.,2015). BiasedSymptomReportingandAntisocialBehaviour 3 The idea – as endorsed by the DSM-5 – punishment, but not mandatory; patients can that antisocial behaviour is associated with opt for incarceration in a penitentiary (and symptom over-reporting is further examined alsooptforthisduringtheirstay).Allpatients in two studies. Unlike other studies in this undergopsychologicalandneuropsychological field, symptom under-reporting (Study 1) and assessment upon admission to establish diag- socialdesirability(Study2)arealsoexamined, noses as laid out in the DSM-5. In addition to because such behaviour is no less deceptive individually tailored treatment for their psy- than symptom over-reporting, and thus – at chopathology, patients are obliged to partake face value – it is equally plausible for it to be ingrouptherapiesaimedatresocialisationand related to antisocial and psychopathic behav- recidivism risk reduction. Patients remain in iour. To explore the idea that the relations treatment until theyare ready to re-enter soci- betweenbiasedsymptomreportingandantiso- etyoruntiltheirprisontermends. cial features are context dependent, a thera- Treatment supervisors provided the peutic forensic setting (Study 1) is contrasted names of patients they deemed fit to partici- withapunitiveforensicsetting(Study2). pate and these patients were then invited to take part in the study. Exclusion criteria (as determined by treatment supervisors) Study1 included insufficient command of the Dutch language, extreme symptoms of drug with- The aim of Study 1 is to investigate the rela- drawal, severe mental instability due to psy- tionship between biased symptom reporting chosis, or deficient mental abilities owing to (i.e. over-reporting and under-reporting) and severe intellectual disability. On these antisocial behaviour (measured by institu- grounds, 25 patients were not approached. tional misbehaviour and ASPD diagnoses) in The majority of these patients were found to a forensic psychiatric context. Given the be unfit to participate because of substance mixed findings in the literature mentioned withdrawalsymptoms,severeintellectualdis- earlier, it was expected that little to no rela- ability, or psychosis. Furthermore, as partici- tionship between antisocial behaviour and pation was voluntary and did not yield any biased symptom reporting would be found. rewards,13eligiblepatientschosenottopar- The therapeutic environment of the psychiat- ticipate.Another3patientsabscondedbefore ric hospital (in which treatment progress behavioural observations were completed, leads to privileges such as furloughs and and1patientchosetoterminatehisparticipa- access to accommodation for recreational tionshortlyafterstartingthefirsttest. activities)createdtheanticipationofahigher In total, 57 male inpatients aged 19 to rate of symptom under-reporting than symp- 54years(MD40.0,SDD9.1)completedthe tom over-reporting. Prior to data collection, study. The mean IQ was 88.2 (SD D 14.6, ethical approval was obtained from the Ethi- range D 61–140; IQ scores are missing for 2 cal Committee of the Faculty of Psychology participants). IQ data were gathered from andNeuroscience,MaastrichtUniversity,and patient records, which contained Wechsler from Radix Forensic Psychiatric Hospital, Adult Intelligence Scale – Fourth Edition Heerlen,TheNetherlands. (WAIS-IV; Wechsler, 2008) protocols. The majority of participants are Caucasian (79%, nD45)andallexceptone(98%,nD56)had Method been diagnosed with one or multiple sub- Participants stance disorders. Furthermore, 28% (n D 16) Participants were recruited from Radix, a had received a diagnosis of other specified medium security forensic psychiatric hospital personality disorder, 16% (n D 9) had been in the Netherlands that admits patients post- diagnosedwithASPD,14%(nD8)hadbeen trial. Confinement there is in lieu of regular diagnosedwithautismspectrumdisorder,7% 4 A.vanImpelenetal. (n D 4) had been diagnosed with attention- my problems under full control’, and an deficit hyperactivity disorder (ADHD), 7% example of a social desirability item is ‘I try (nD4)hadbeen diagnosed withschizophre- to help everybody who has problems’. nia, and 19% (n D 11) received no diagnosis Endorsement of supernormality and social otherthansubstancedisorder. desirabilityitemsissummedsoastoobtaina totalSSscore.Whiletheinternalconsistency (Cronbach’salphaD.86)andtest–retestreli- Measures ability(rD.90)oftheSSaresatisfactory,the TheStructuredInventoryofMalingeredSymp- diagnostic accuracy indices are meagre, with tomatology(SIMS). TheSIMS(Smith&Bur- the sensitivity and specificity being .74 and ger,1997;seeMerckelbach&Smith,2003for .42 for a >14 cut-off score, .58 and .67 for a the Dutch translation) is a symptom validity >17 cut-off score,and .28 and .93for a>21 test that assesses a broad spectrum of feigned cut-off score, respectively (Cima et al., andexaggeratedsymptoms.TheSIMSconsists 2003). However,the SS does possess moder- of 75 true–false items, which constitute five ate predictive validity, as undergraduate stu- subscales that target feigned depression, psy- dents instructed to imagine that they were chosis, neurologic impairment, memory dys- offenders who opted for parole and who function, and low intelligence. The items therefore engaged in faking good exhibited mostlyrefertobizarreexperiencesandatypical statisticallyhigherscorescomparedtocontrol symptoms such as ‘I have difficulty recognis- individuals(Cimaetal.,2003). ing written and spoken words’ and ‘When I can’tremembersomething,hintsdonothelp’. The Levenson Self-Report Psychopathy Scale Thenumberofendorsedsymptomsissummed (LSRPS). The LSRPS (Levenson, Kiehl, & so as to obtain a total SIMS score. The SIMS Fitzpatrick, 1995) is a 26-item self-report does not require a high reading level (i.e. instrumentthatassessestraitsassociatedwitha Flesch–Kincaid Scale 5.3 suffices; Smith, callous and manipulative orientation towards 2008).TheinternalconsistencyoftheSIMSis others (i.e. primary psychopathy) and with a reasonable (Cronbach’s alpha coefficients of disinhibitedandantisociallifestyle(i.e.second- .72foundbyMerckelbach&Smith,2003,and ary psychopathy). Items are scored on four- .92 to .94 by Rogers, Robinson, & Gillard, point Likert scales (where 1 D strongly dis- 2014). Both studieswith experimental simula- agree and 4 D strongly agree). A total score tors and studies with identified malingerers (i. that is reflective of psychopathic traits can be e. known-groups studies) haveyieldedaccept- calculated by recoding some items and then ablediagnosticaccuracyparameters,withsen- summing all scores. Total scores of >57 are sitivity circling around .91 and specificity considered to be ‘high’ (Brinkley, Schmitt, around.65foracut-offof>16(foradetailed Smith, & Newman, 2001). A representative overview,seevanImpelen,Merckelbach,Jeli- itemfromtheprimarypsychopathysubscaleis cic,&Merten,2014). ‘[i]ntoday’sworld,Ifeeljustifiedindoingany- thingIcangetawaywithtosucceed’,whereas The Supernormality Scale (SS). The SS an illustrative item from the secondary psy- (Cimaetal.,2003)isaself-reportinstrument chopathyscaleis‘Ihavebeeninalotofshout- thathasbeendevelopedasaresearchtoolfor ing matches with other people’. While the assessing symptom under-reporting. It con- LSRPS was originally designed to assess psy- sistsof37true–falseitems,ofwhich21items chopathic traits in non-institutionalised sam- comprise a supernormality subscale (measur- ples,ithasbeenemployedsuccessfullyinlarge ing minimisation of mild psychopathological forensic samples (Brinkley et al., 2001; Wal- phenomena), 11 items comprise a social ters, Brinkley, Magaletta, & Diamond, 2008). desirability subscale, and 5 items are bogus. The internal consistency of the total and Anillustrativesupernormalityitemis‘Ihave primary psychopathy scale is adequate BiasedSymptomReportingandAntisocialBehaviour 5 (Cronbach’salphaD.82–.84),whereasthatof Malingering Scale (Schretlen & Arkowitz, the secondary psychopathy scale is moderate 1990), which are cognitive paper-and-pencil (Cronbach’salphaD .63–.68;Levensonetal., tasks that measure underperformance. After 1995;Lynam,Whiteside,&Jones,1999). participants had completed the test battery, their engagement in institutional misbeha- The Social Dysfunction and Aggression viour was monitored for a period of six Scale-11 (SDAS-11). The SDAS-11 (Wis- weeks. The monitoring of participants was tedt et al., 1990) is an 11-item behavioural achieved through the close examination of observation scale that was developed to mea- patient records, which were maintained on a sure social dysfunction and aggression in psy- daily basis by nursing staff, therapists, psy- chiatric inpatients. The SDAS-11 is scored chologists,physicians,andtreatmentsupervi- over a longer time interval, with one-week sors. The patient records that were used for intervals between successive ratings. It con- the present study contain reports of daily sists of 9 items covering outward aggression activities and detailed accounts of clinically and social dysfunction (e.g. irritability, nega- relevant activities and behaviour, such as tivism, verbal and physical aggression) and 2 social functioning, treatment progress, and items covering inward aggression (i.e. self- physical and emotional well-being. As such, harm), with each item including a five-point thepatientrecordscontainampleinformation scoringscale(rangingfrom0Dnotpresentto for completing the SDAS-11 items. For each 4 D severe). The outward and inward items participant,thescoresonthetestbatterywere arenotinter-correlatedandtheinternalconsis- calculated only after scores on the SDAS-11 tency is acceptable (Cronbach’s alpha D .79; had been obtained, thus reducing experi- Wistedt et al., 1990). In the current study, menter bias during the evaluation of patient SDAS-11 items were used to evaluate the recordswithSDAS-11items. dailynurseobservationrecordsofeachpartici- pant. More specifically, for each participant, Results all records of a six-week period were selected andscoredintermsofindicationsforthepres- Table 1 summarises the mean scores on the ence of SDAS-11 items. To explore the reli- psychometric instruments and also gives the ability of this procedure, a random set of ten proportion of patients who scored above cut- one-week records from 10 patients were off points, as well as the prevalence of the selected and evaluated by the first author and most frequent diagnoses among these another rater who was alsoblind to the symp- patients. As can be seen, symptom under- tom validity status associated with each reporting(asindexedbytheSS)ismorethan record. The Spearman rank order correlation twiceasprevalentassymptomover-reporting betweenthetworatersis.79. (asindexedbytheSIMS). The most frequent diagnoses among par- ticipantswhofailedtheSIMScut-offscore(n Procedure D 5, 9%) are an IQ of <75 (60%) and other Seated in a small therapy room on their own specified personality disorders (OSPDs; ward,participantsfirstgavewritteninformed 40%). Among participants who failed the SS consentthatwasalsoverballycommunicated cut-off score (n D 13, 23%), the most fre- tothem.Next,participantscompletedthetest quentdiagnoses are OSPDs (38%)andan IQ battery, which included – in counterbalanced of <75 (23%). Furthermore, two participants order–theSIMS,theSS,andforasubsample produced a significantly outlying score (>2 (n D 25) also the LSRPS. The test battery SDs) on the LSRPS; they were diagnosed also included an instrument that is not with ASPD and autism spectrum disorder, addressed in the current study: the Vocabu- respectively. The group of participants who lary and Abstraction subtests of the scoredbeyond1.5SDsontheSDAS-11(nD 6 A.vanImpelenetal. Table1. Summaryofmeans,SDs,95%confidenceintervalsandprevalenceratesofdiagnosesandcut- offfailuresintheforensicpatientsample(nD57). Percentageofmost frequentdiagnoses Percentageexceeding ofpatients M(SD) 95%CI cut-off scoring>cut-off SIMS 8.6(6.6) [6.8,10.3] 9%(5outof57)>16 60%IQ<75,40%OSPD SS 16.4(6.5) [14.7,18.2] 23%(13outof57)>21 38%OSPD,23%IQ<75 LSRPS 52.6(9.6) [48.6,56.6] 28%(7outof25)>57 Nodiagnoseswithacount>1 SDAS-11 17.1(12.5) [13.8,20.4] 16%(8outof57)>1.5SDs 38%IQ<75,25%OSPD Note:LSRPSDLevensonSelf-ReportPsychopathyscale;OSPDDotherspecifiedpersonalitydisorder;SDAS-11D SocialDysfunctionandAggressionScale-11;SIMSDStructuredInventoryofMalingeredSymptomatology;SSD SupernormalityScale. 8, 14%) is diverse with regard to diagnoses: psychopathictraits;LSRPSscores>57,insti- three participants (38%) had an IQ of <75 tutional misbehaviour; SDAS-11 scores > andtwo(25%)hadanOSPDdiagnosis.There 1.5 SDs above the mean, and ASPD diagno- is no overlap between the group of partici- ses) and biased symptom reports (i.e. symp- pants who exceeded the cut-off score of the tom over-reporting; SIMS scores >16, and SIMS and the group who exceeded the cut- symptom under-reporting; SS scores >21). offscoresoftheSSortheLSRPS.Ofthethir- Fisher’sexacttestsindicatethatbiasedsymp- teen participants who scored above the cut- tom reporting is not associated with psycho- off on the SS, three (23%) have SDAS-11 pathic traits, institutional misbehaviour, or scores of 1.5 SDs above the mean and two ASPDdiagnoses(allps>.05). have an LSRPS score that surpasses the cut- As another approach to data analysis, offof>57.Interestingly,themeanSDAS-11 Pearsonproduct–momentcorrelationsamong scoreofASPDpatientsdoesnotdifferstatis- the various measures were calculated ticallyfromthatoftheotherpatients(15.4vs (Table 2). Neither self-reported psychopathic 17.3), t(55) D 0.3, p D .75, a result that traits(LSRPS)norinstitutionalmisbehaviour squareswiththerecentfindingofEdens,Kel- (SDAS-11)were found tobestronglyrelated ley, Lilienfeld, Skeem, and Douglas (2015) tosymptomover-reporting(SIMS)orunder-- thatASPDhasnopredictivevalueforinstitu- reporting (SS). Institutional misbehaviour tionalmisconduct. (SDAS-11) was found to be unrelated to To examine the relation between antiso- symptom over-reporting (SIMS) or under- cial features and biased symptom reporting, reporting(SS),whereasself-reportedpsycho- binary contingency tables were computedfor pathic traits (LSRPS) seem to be moderately all antisocial behaviour indices (i.e. associated with symptom over-reporting Table2. Pearsonproduct–momentcorrelationsand95%confidenceintervalsfortheStudy1data. Measure 1SIMS 2SS 3LSRPS 2SS ¡.28(cid:1)[¡.50,¡.02] – 3LSRPSᵃ .31[¡.09,.63] ¡.10[¡.30,.48] – 4SDAS-11 ¡.04[¡.30,.22] .00[¡.26,.26] .27[¡14,.60] Note:(cid:1)p<.05,two-tailed;ᵃnD25.LSRPSDLevensonSelf-ReportPsychopathyScale;SDAS-11DSocialDysfunc- tionandAggressionScale-11;SIMSDStructuredInventoryofMalingeredSymptomatology;SSDSupernormality Scale. BiasedSymptomReportingandAntisocialBehaviour 7 Table3. Numbersofforensicpatients(nD57)withincreasing,constant,ordecreasingantisocialbehav- iourwhoover-reportorunder-reportsymptoms. Over-reporting(SIMS) Under-reporting(SS) SDAS-11 Belowcut-off((cid:2)16) Abovecut-off(>16) Belowcut-off((cid:2)21) Abovecut-off(>21) Increasing 21(37%) 2(4%) 18(32%) 5(9%) Stable 24(42%) 2(4%) 20(35%) 6(11%) Decreasing 7(12%) 1(2%) 6(11%) 2(4%) Note:SDAS-11DSocialDysfunctionandAggressionScale-11;SIMSDStructuredInventoryofMalingeredSymp- tomatology;SSDSupernormalityScale. (SIMS), but not under-reporting (SS). Age is Discussion not related to any of the measures; IQ is Theprevalenceofsymptomover-reportingin related only to symptom under-reporting, this sample of forensic psychiatric inpatients with those with higher IQs predisposed is relatively low (9%) compared with esti- towards less under-reporting, r D ¡.30 mates that can be found in the literature (cf. [¡.52,¡.04],pD.03,two-tailed. 19%:Mittenberg,Patton,Canyock,&Condit, Consistentwiththecorrelationalanalyses, 2002; 32%: Pollock, Quigley, Norley, & multiple linear regression analyses indicate Bashford,1997).Theparticipantsinthepres- that neither self-reported psychopathic traits ent study were recruited from a forensic psy- (LSRPS) nor institutional misbehaviour chiatric hospital where patients are admitted (SDAS-11) are predictive of symptom over- oncetheirsentenceshavebeenpassed,andin reporting (SIMS), F(2, 22) D 1.70, p D .21, which patients have relatively few apparent orsymptomunder-reporting(SS),F(2,22)D externalincentivestoover-reportsymptoms– 0.40, p D .67. Analyses including the scores in fact, doing so may even result in delayed onthesubscalesoftheSIMS,SS,andLSRPS furloughs and prolonged stays. This might did not yield additional information. In sum, explain why symptom under-reporting was no association was found between institu- more than twice as prevalent as symptom tionalmisbehaviourandeitherformofbiased over-reportinginthepresentsample (23%vs symptom reporting, nor is there a relation 9%). Additionally, the prevalence of symp- between biased symptom reporting and tom over-reporting may have been low ASPD or self-reported psychopathy (yet the because of selection bias: treatment supervi- latterisbasedonnD25). sors prohibited the inclusion of patients they Next, groups were formed based on the deemed too disordered to participate. It may temporal trends of the SDAS-11 scores; one bethataportionofthesepatientsexaggerated group had scores that increased over time their pathology (and would have over- (nD23),onegrouphadscoresthatremained reportedsymptomshadtheyparticipated). relativelystable(nD26),andonegrouphad TheprevalenceofASPD(16%)andantiso- scores that decreased over time (n D 8). The cial behaviour is low as well. A large portion threegroupswerethencomparedwithregard of patients’ SDAS-11 scores are explained by to biased symptom-reporting (i.e. frequency irritability, negativism, mild resentment, and of individuals scoring above the SIMS or SS moderateverbalaggression. Thus,participants cut-offs).Table3showsthepatterns.Fisher’s engaged almost exclusively in mild disruptive exact tests yielded no significant results (all behaviour; none engaged in serious physical ps > .05), which implies that institutional violence, self-harm, or severe verbal aggres- misbehaviour is in no way related to symp- sion.Therelativeabsenceofgravelydisruptive tomover-reportingorunder-reporting. behaviourislikelyduetoseveralfactors.First, 8 A.vanImpelenetal. it may have to do with the focus on treatment and invited to participate in two test sessions and the consequently comprehensive and con- without compensation. Insufficient literacy stantimplementationofazerotolerancepolicy and command of the German language are onallwards,whichhaveastafftopatientratio the only exclusion criteria. A total of 65 ofatleast1:6(usually1:4).Second,itmaybe inmates agreed to participate and completed relatedtotheconsiderableweightoftheconse- the first session, but only 45 of those com- quences of misbehaviour, which typically pleted the second session, which took place include delayed or revoked furloughs, pro- two to three weeks later. Reasons for drop- longed stays, or – in severe cases – relocation ping out included completion of the prison toapenalinstitution. sentence, relocation to another facility, and lack of interest in the second session. The majority of the final sample (n D 45) were Study2 sentenced prisoners, and 4 (9%) were on This study examines the relation between remand.Themeanagewas20.7years(SDD biased symptom reporting and antisocial 1.7,rangeD18–24). behaviour in a punitive forensic setting.Addi- tionally,severaltypesofblameattributionand excuse-making are assessed. Niesten et al. Measures (2015)reportaninterestingdifferencebetween Antisocialanddelinquentbehaviourwasmea- forensicpsychiatricpatientsandprisonerswith sured with several proxies: length of prison respect to symptom over-reporting and under- sentences(inyears),numberofincurreddisci- reporting. More specifically, they found that plinary actions (coded as a continuous vari- both types of distorted symptom reporting are able),andclassificationas‘intensiveoffender’ higherinthelattergroup,presumablybecause (yes/no). Intensiveoffender(Intensivta€ter)isa the incentives to distort symptoms are higher term used in Germany to designate juveniles inthatcontext.Withthisinmind,moresymp- whose delinquency is serious and repetitive. tom over-reporting was expected in the puni- Althoughtherearenoformaldefinitionsorcri- tive setting of Study 2 than in the therapeutic teria to establish intensive offending, the term setting of Study 1, yet it was also predicted iscommonplaceintheGermanjusticesystem. that the relationship between biased symptom The SIMS and the LSRPS were used (see reporting and antisocial behaviour would be Study 1 for details), along with two other similarly small. Cima et al. (2003) observe instruments. The test battery also included a that symptom under-reporting in a forensic measure of symptom overreporting that was sample is related to the tendency to blame not used in the analyses below (the recently external conditions or others for their crimes. developed Self-Report Symptom Inventory; To extend this work, it was decided to test SRSI;Merten,Merckelbach,Giger,&Stevens, whether symptom over-reporting is associated 2016). with excuse-making and blame attribution to mentaldisorders.Approvalwasobtainedfrom the standing Ethical Committee of the faculty The Social Desirability Scale-5 (SDS-5). of Psychology and Neuroscience, Maastricht The SDS-5 consists of five modified items University, and the Youth Prison of Berlin from the Social Desirability Scale-17 (SDS- (JugendstrafanstaltBerlin),Germany. 17; St€ober, 2001). The internal consistency (Cronbach’salphaD.80)andtest–retestreli- ability (r D .82) of the SDS-17 are adequate. Method The SDS-5 was embedded in the LSRPS. Participants Therefore, the original true/false format is Inmatesofanall-maleyouthprisoninBerlin replacedwithafour-pointscale(where1DI wereprofferedabriefdescriptionofthestudy donotatallagreeand4DIfullyagree). BiasedSymptomReportingandAntisocialBehaviour 9 The Revised Gudjonsson Blame Attribution SIMS, the BAI, and the LSRPS with the Inventory (BAI). The BAI (Gudjonsson & SDS-5 items. Participants were told that the Singh, 1989; for the German translation, see instrumentsmeasuredpersonalitycharacteris- Cima et al., 2006) contains 42 items that tap ticsandpsychologicalproblems. into three independent dimensions of blame attributionforcriminaloffences:externalattri- bution (i.e. blaming transgressions on social Results environments, victims, or society; Cronbach’s Table 4 presents means scores of the sample alpha D .77), mental-element attribution (i.e. on the various measures, as well as the pro- placing blame on mental disorders or insuffi- portion of prisoners whose scores exceed the cient self-control; Cronbach’s alpha D .79), associatedcut-offs.Ascanbeseen,13%(nD andguilt-feelingattribution(i.e.feelingremorse 6) of the participants failed the SIMS, which or regret about offences; Cronbach’s alpha D is only slightly higher than the failure rate in .81). Items consist of first-person statements Study 1 (9%). More than half of the sample that are evaluated on a five-point scale (where scored above the cut-off on the LSRPS 0 D I do not at all agree and 4 D I fully (56%), which is considerably higher than the agree). Representative items include ‘I did not proportion with extreme LSRPS scores in deservetobecaughtforthecrimeIcommitted’ Study1(28%). (external attribution), ‘I would certainly not Table 5 displays the correlations among have committed the crime I did if I had been thevariousmeasures.Symptomover-reporting mentally well’ (mental-element attribution), (SIMS)correlatespositivelywithself-reported and‘IhavenoseriousregretsaboutwhatIdid’ psychopathictraitsasmeasuredbytheLSRPS (guilt-feelingattribution). and negatively with social desirability as indexed by the SDS-5. Symptom over-report- ingisalsorelatedtoblameattributiontoexter- Procedure nal factors such as social environments, Two sessions were undertaken, the first con- victims,orsociety(BAIExternal)andtomen- taining the informed consent form followed tal disorders (BAI Mental). However, symp- by administration of the SRSI. The second tom over-reporting is not statistically related session(whichtookplacetwotothreeweeks to behavioural proxies of antisocial behaviour later) consisted of administration of the (i.e. sentence length, number of incurred Table4. Summaryofmeans,SDs,95%confidenceintervalsandprevalenceratesofcut-offfailuresinthe forensicpunitivesample(nD45). M(SD) 95%CI Percentageexceedingcut-off SIMS 10.6(5.2) [9.1,12.2] 13%(6outof45)>16 SDS-5 16.7(2.7) [15.9,17.5] N/A LSRPS 58.9(9.3) [56.1,61.7] 56%(25outof45)>57 BAIExternal 29.5(8.5) [26.9,32.0] N/A BAIMental 25.0(6.6) [23.0,27.0] N/A BAIGuilt 54.6(11.7) [51.1,58.2] N/A Prisonterm(years) 2.5(1.4) [2.1,2.9] N/A Punitiveactions 0.5(0.9) [0.3,0.8] N/A Intensiveoffender N/A N/A 22%(10outof45) Note:BAIDRevisedGudjonssonBlameAttributionInventory;LSRPSDLevensonSelf-ReportPsychopathyScale; PunitiveactionsDrelativenumberofpunitiveactionstakenagainstparticipants;SDS-5DSocialDesirabilityScale-5; SIMSDStructuredInventoryofMalingeredSymptomatology.

Description:
Behaviour in Forensic Samples: A Weak Link, Psychiatry, Psychology and Law. To link to this article: victims, or society (BAI External) and to men- tal disorders (BAI Mental) .. A comparison of actual test scores, however, reveals
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.