PsychExchange.co.uk Shared Resource

Cognition BiologyUpbringing

*Reductionist

*Difficult to investigate cause and effect

*Difficult to isolate confounding variables

*Usefulness; cannot explain all crimes e.g. crimes of passion

*More holistic in its approach *Deterministic

*Application to everyday life; interventions

*By its very nature must use longitudinal design

*Application; Cognitive skills Retraining (CBT)

*Application; potentially worrying (Minority report)

*Used in crime prevention to alter criminal appraisals of opportunities

*Can take account of both biology and cognition

*Individual diffs. Not all P with low front lobe=crim

*Free-will

*Cannot ‘see’ thought processes. Interviews/qres

*Physiological measures:Ex.bias, E.V.

Farrington Sutherland (orig.1939, pub 1947) Wikstrom & Tajfel

*Theoretical rather than empirically supported.

*Useful as formed basis of S.L.T.

*Application; Sutherland argues may be possible to create a formula…

*Does not explain solo crimes

*opposes Phys., emphasising social aspects.

*Nature/Nurture

*Good sized sample (411)

*Longitudinal design, low attrition rate

*Andro- and ethno-centric

*Has been running for 49yrs so far

*Intergenerational transmission

*Intervention suggestions (App./usefulness)

*Both qual.& quant. Data

*Self-report

*Dispositional/Situational

*Cross-sectional design

*Good sized sample (2000)

*App: Identified explanatory factors and protective factors.

*Stimulated further research into influencing factors.

*Strength/limitation of research

Gudjohnsson & Bownes Yochelson & Samenow Kohlberg 1963

*Took place over 14yrs

*Androcentric 255males

*Sample and attrition issues

*Used Psychodynamic approach

*App: Only 9 genuinely changed

*Validity: P’s lied

*No Control group as evolved from clinical cases

*Built on work of Piaget

*Sample 58 boys Chicago

*Some followed longitudinally

*2hour interviews, dilemmas e.v.?

*Follow up studies world-wide and on females too

*Reviewed theory, no 6

*Gilligan criticised

*App.treatment progs

*Social cognition

*42 item self-report GBAI

*80ps Andro/Ethno English results differed

*Criminals guilty of different crimes show differences in levels of guilt and attributions.

*App. May help to retrain attributions. LH, SSB

*Qual/Quant. data

*Correlational study: CDNIC

*Life-expectancy -ve corr with homocide (-0.88)

*-ve corr. School attendance and life expectancy. P<0.001

*Reductionist

*Cannot assess cause and effect

*Appn;

*Individual Differences

Raine Brunner Daly & Wilson

*Review article

*Not strictly reductionist or deterministic.

*Birth complications, frontal lobes, resting heart rate.

*App.reversal of predisposing factors.

*Validity based on quality of reviewed research

*Ethics.

*Five Dutch males eth/andro

*External validity

*Urine analysis

*Reductionist. Not all male members of family were violent despite MAOA def.

*Usefulness: rare deficiency

*Approaches

*Application…ethics

Interviewing witnesses Offender ProfilingInterviewing suspects

*Reductionist

*Reliability: Top-down, bottom-up, UK inconsistencies

*Difficult to ethically create high E.V. studies

*Difficult to ethically create high E.V. studies

*Appn: Potentially very useful

*Application; potentially brilliant..Copson found not

*Samples usually not representative

*Individual diffs. Not all crims are predictable

*Often use lab studies

*Validity: How do you measure success

*Appn: Potentially very useful

*Often use lab studies

*Samples usually not representative

Bruce et al Loftus et al Fisher et al

*Lab study.

*Sample small &not representative, 18-31, 36 students paid or credited (dem.chars?)

*Application; Memory can be adversely influenced when a weapon is present.

*Poor E.V., Gun=fear!

*Self-report measures (qre,confidence), eye-fixation measures.

*Deception “Proactive interference”

*Lab study

*Small samples (30,48)

*Ethnocentric (Stirling Uni)

*Paid £2 to sort composites

*Ind.measures Full,ext.,int.

*Appn. Ext.features more important than int. in ID

*Effect still holds for familiar faces

*E.V. ID of crim is really important

*Methodology: lab study, foils

*Field study

*16 US detectives

*Sound methodology, CI trained & Control group, ratings by ‘blind’ to CI/SI.

*App: CIT effective in field and is used by police & therapists.

*Good E.V. Previous lab-based studies found N.S.D.

*Still under development

Mann et al Inbau et al Gudjohnsson et al 1990

*Interrogation/Miranda rights

*Procedure used in USA =ethnocentric

*App…you can’t. P.A.C.E.

*’Method’ may produce results which lack validity

*Inbau says effective, vulnerable=false confessions.

*Ethics: Deception, P.O.P.

*obviously right to withdraw

*Case study

*17yr old youth

*P.A.C.E. would prevent this

*interrogated for 14hrs

*Appn. Gudjohnnson Suggestibility scale identified vulnerability

*E.P.I. S.E., I.Q.94

*Coerced Compliant (see Kassin &Wrightsman 1985, Horselenberg)

*E.V.=real life

*Field experiment

*99 Kent police 75m24f

*Qre detecting lie experiencing

*54 Video clips of 14 suspects.

*Forced choice, confidence and cues

*Qual.&Quant.

*More accurate than chance.

*Corr. Experience & accuracy (SEE BOOK)

*No control group (P.O.P)

*E.V.? Story cues missed.

*Case study*Appn: Geographic

profiling*Usefulness:2000/1502/

3rd led to surveillance

*Reductionist? Staging

*Effectiveness: Profiling point accurate but critics state only the right remembered

Canter et al Canter & Heritage Canter-John Duffy

*100cases content analysiscontent analysis, multi-dimensional scaling

*serial killers 3rd murder categorised using Crime Classification Manual.

*Method: organised component in all 3rd murder not caught

*Disorganised more common, Organised hid body, sex activity

*Validity based on quality of police reports

*Dichotomy ignores ind.diff

*Content analysis 66offences, 27 offenders

*Small space analysis (corr.based)

*Appn.5-factor theory may help identify whether same offender responsible M.O. or whether M.O. changing.

*Sex offences crime reports (P.O.P.=police responsibility)

Persuading a jury Reaching a verdictWitness Appeal

*déjà vu?

*Low in E.V.,Validity

*Generalising from non-court-room studies e.g. Stoner, Asch

*Difficult to ethically create high E.V. studies

*Difficult to ethically create high E.V. studies. Only mock and Shadow juries allowed

*Appn: Potentially very useful *Application; potentially

brilliant.*Samples usually not representative

*Individual diffs. Not all juries are predictable

*Often use lab studies

*Could use deception more

*Appn: Potentially very useful

*Often use lab studies

*’Appeal’ can be subjective

*Samples usually not representative

*Attractiveness results mixed (see Bull & McAlpine)

Castellow Penrod & Cutler Ross

*Lab study, mock trial, ind.measures, counterbalanced order of presentation.

*Sample 538 jury-eligible students credits (dem.chars?)

*Witness confidence only one of 10 variables with p<0.05, in later research (x9!) confidence & accuracy however p>0.05 N.S.D.

*Application; Witnesses with belief are more convincing but not necessarily more accurate.

*Poor E.V.

*Lab study, independent measures

*ψ Student sample (71m,74f)

*Ethnocentric (East Carolina Uni)

*Gained extra credit

*Photos previously rated by panel for attractiveness

*P’s rated photo as guilty and rated personality on 11bipolar forced-choice scales

Appn. “Halo effect” (Asch) still present, may need to warn jury, defendant dress well

*Mock jury

*E.V. judging guilt by looking at a photo and a case booklet?

*Mock trial based on real transcript. Actors played roles

*3 Ind.measures. Screen, video-link, present (N.S.D.)

*Jury warned in non-present conditions not to take non-presence as sign of guilt.

*300 students 150m,150f

*P’s rated child on credibility of aspects of story, rated defendant on similar.

*App: Gender diff. of guilt/child credibility verdicts may be important in peremptory challenges/voire dire

*Poor E.V. real transcript does not make it real-life.

*Poor P.o.p but how else can we investigate this?

Pickel Cutler et al

*Not pure Mock trial, Ind.measures, lab.experiment, 2 groups and control

*236 Bali State ψ students

*Course requirement

*Recording of a case, qre-verdict, estimation of guilt, likert scale of how much inadmissable evidence influenced verdict.

*Appn. Telling jurors to ignore evidence tends to make them focus on it!

*538 students extra credits

*Viewed Videotaped robbery in groups of 2-8; Qre:verdict, memory test, confidence scale.

*4 IV’s witness confidence, conditions, descriptive v numerical, expert opinion.

*App 85% of testimony recall was good.Expert testimony helped witnesses, could help prevent miscarriages of justice

*Lab.study, not pure mock trial

*130 Students from two US Unis, paid, allocated to one of 4 ind.groups.(def.story/witness order X pros.story/witness order)

*Tape recording of a case, P’s rated Guilty or not and confidence on 5 point likert scale. On own, no delibs.

*Appn. Story order most persuasive and fills jury with more confidence in their conviction…or not.

*Methodology: defence case less persuasive due to nature of source material.

*Real source, still poor e.v.

Pennington & Hastie

*Lab.experiment, not based on juries, confederates

*In no pressure situation only 1/35 gave wrong answer, in conformity condition 32%

*Appn people conform to be right or to be part of a group, they may need to be made aware of this prior to trial.

*Ethics: deception

*E.V. Jury task more complex than line length

AschMoscoviciHastie

*Review paper published in the American Bar Association Journal.

*Only as good as papers it reviews…based predominantly on mock/shadow juries

*References removed as book was to be published

*Appn. ?

Paper itself has recommendations which are useful in other areas

*6 participants 2 confederates, colour perception tasks…relates how to a jury?

Appn. Consistency of minority is persuasive (see Nemeth & Wachtler for further support). Consistency equates to confidence…people follow confidence

Nemeth & Wachtler

*Lab.mock trial with confederates.

*Groups of 5…jury size?

*Compensation decision

*Choosing head seat, history of influence & consistency persuasive

Imprisonment Treatment ProgrammesAlternatives to Imprisonment

*Hawthorne effect

* Should have control groups, matched pairs

*Risk of Hawthorne effect*Sometimes diff. for independent researchers to gain access

*Appn: Potentially very useful, help convicts, reduce recidivism, save money

*Effectiveness...how measured?

*Studies often dated (Tosh 1982; Bartol & Bartol 1994; Dooley 1990)

*Individual diffs. *Subject to demand chars

*Could use deception more

*Appn: Potentially very useful RRRID. It’s expensive, can we make it effective?

*Validity

*Politics/usefulness

*Effectiveness…how measured?

*Should have control groups, matched pairs

*Appn: Potentially very useful, help convicts, reduce recidivism, save money

Gillis & Nafekh Dooley Haney & Zimbardo

*Content analysis.

*442 deaths E&W, ’72-87

*Checklist Social, Psychiatric & forensic history (qual & Quant. Data)

*Application; Consistency needed in classification of prison deaths.*Psychological assessment/monitoring of vulnerable needed

*Overcrowding (supported by Calhoun, Haney & Zimbardo) & remand increase risk factors

*Prison population has almost doubled since this study

*Content analysis

*23,525 Inds (Jan 98-05) Good sample

*Ethnocentric (Canada)

*Matched pairs (gender, risk level, release yr, sentence length, family/marital situation, substance abuse + others)

*Appn. Integration increased S&N rates 70%v55%

*Strong methodology

* Ψ as science: Supports Azjen & Fishbein

*Review paper. Summarises changes in US prison system and makes suggestions for future.

*Empirical research more useful?

*1998 is now 12 years ago

*US system very different=ethnocentric

*Experimenter bias?

*P’s rated child on credibility of aspects of story, rated defendant on similar.

*App: Makes specific suggestions as to how US system should change including

*prison-specific Psychological assessments (sit. Not relevant in prisons).

*Reform needs to come from outside system

*Greater use of Psychological input

*Qual., Quant. & anecdotal evidence

Mair & May (Probation) Eberhardt (Deathworthy)2006

*Survey Piloted 7offices 24p’s (ceiling/floor)

*3299 random from 22offices E&W. 40% failed to turn up Sample biased (less London)

*Corrective weighting.. Methodological shortcut

*Interviewers independent

*Data lacked open qns

*Demand chars

*Appn. Reduce recidivism, help crims, save money

*Lab.study.

*Ethnocentric (thankfully?)

*Database analysis ‘79-99, 44 photos of black murderer of white victim.

*51 raters from Stanford Uni. Rated photos (b&w 4secs) from 1-11, 11=stereotypical.

*ANCOVA found degree of stereotyp. Blackness most sig. variable. Followup study found no diff. when victim was black.

*App Racism in CJS

*Internet search!!!424 articles

*36 studies

*Appn. Reduction in re-offending rates for violent and property crime, Quicker recovery from PTSD for victim

*Methodology: Review article issues, plus internet appearance less reliable than peer-reviewed journals.

*

Sherman & Strang (RJ)

*Field study ’96-00

*180f-14 non-completers

*Control group 540

*Expected 2yr conviction rates calculated, matched on risk level.*Actual reconviction rates calculated

1&2yrs* treatment v none NSD*R&R treatment sig worse*Reliability of methodology*Treatments not effective for women*F cognition differs from M (all those

strappy shoes)*No Hawthorne effect here!

Appn Just receiving a treatment is not enough to ensure improvement.

Cann (Cog)Wheatley (ears)Ireland (anger)

*Field study, quasi experiment 50 AM 37 not

*Qual. & quant. Data (Cog.Beh.interview, WBC completed by prison officers, Self-report qre, completed before and after treatment (or not treatment)

*AM group decreased, Hawthorne effect? Placebo?Dem.Chars?

*Effectiveness? No reconviction data

Appn. Short term gains? 8% got worse!

*350 H.Sec. prisoners. *Exp.grp=Acup+S.C.prog, control grp=S.C.

*Qual&quant data: Social improvements, drug reductions

*Hawthorne?Placebo?Dem.Chars? Does it matter?Appn…

*Effectiveness? No reconviction data

Gates*Genuine v Sham database of trials.(review paper)

*1433 good sample but…

*Problems exist with review papers measuring like with like. Variations in needle placement

*Psych. As a science