The Risk of Making False Confessions: The Role of Developmental Disorders,Conduct Disorder,Psychiatric Symptoms,and Compliance

Abstract

Objective: To investigate salient predictors of self-reported false confessions. The conditions included mild intellectual disability (ID), autism spectrum disorder (ASD), ADHD, conduct disorder (CD), psychiatric symptoms, and compliance. Method: The sample consisted of 386 male Scottish prisoners, who completed a diagnostic ADHD interview, psychometric tests, and a questionnaire about false confessions they had given to police over their lifetime. Results: There was a high rate of reported false confessions (33.4%); the reason for the majority (62.2%) being to “cover up for somebody else.” CD, ADHD, psychological distress and psychiatric symptoms, and compliance were all significant predictors of false confession. When examining the influence of underlying domains of ADHD, it was found that hyperactivity/impulsivity drove the association with false confession after adjusting for coexisting CD. Conclusion: CD and ADHD were the two most powerful predictors of false confessions, with CD mediating some of the effects of ADHD.

Keywords

false confession developmental disorders compliance

Background

Studies of miscarriage of justice have shown that between 14% and 25% involve false confessions (Drizin & Leo, 2004). In terms of psychological vulnerability, a common theme across studies is that many of the exonerated people, who falsely confessed, had intellectual disability (ID) or mental illness, although authors typically do not present systematic evidence for this claim (Gudjonsson, 2018). Perske (2008) provides an important review of 53 persons with intellectual/developmental disability, including those falling in the “mild” and “borderline” ranges, who had given a false confession to a serious offense and been legally exonerated.

Lower IQ

Studies of police detainees and prisoners have shown them to have a mean Full Scale IQ of about 83 (Gudjonsson, Clare, Rutter, & Pearse, 1993; Hayes, Shackell, & Mottram Lancaster, 2007; Viljoen, Klaver, & Roesch, 2005). It remains unclear whether those who make a false confession are more intellectually disadvantaged than other police detainees or prisoners. This is an important research question. The current evidence for answering this question is somewhat contradictory.

In Iceland, Sigurdsson and Gudjonsson (1996a) found no significant difference in nonverbal IQ between “false confessors” and other prisoners. This study is particularly important in that it included 96% of all prisoners approached to participate (Sigurdsson, 1998), making it easier to generalize from the findings.

Gudjonsson (1990) found that alleged “false confessors” referred for a psychological assessment had a significantly lower Full Scale IQ than other forensic referrals, but this may have been due to a selection bias of suspects referred for an assessment in cases of disputed confessions (Gudjonsson, 2018). In a study of 40 DNA exonerated men who had falsely confessed in the United States, 35% were reported to have ID or were in the borderline range (Garrett, 2011). Schatz (2018) estimated a rate of 25.7% in his analysis of 245 exonerated false confession cases.

In the United Kingdom, Gudjonsson (2010) found a lower rate in a review of 34 cases of disputed confessions where convictions were overturned on appeal. IQ scores were available for 36 (78%) of 46 appellants (Gudjonsson, 2018). The mean IQ for this group was 89 (SD = 14.6, range = 51-113). Only one had an IQ below 70% and 16.7% had scores of 75 or below (i.e., bottom 5%) of the general population. Eleven (30.6%) had an IQ score of below 80 (i.e., either ID or borderline).

There may be other psychological vulnerabilities that potentially increase the risk of a false confession to police, because of impaired capacity to cope effectively with detention and interrogation and make informed decisions (Gudjonsson, 2003, 2018) These include other neurodevelopmental conditions, psychiatric symptoms, and personality factors (e.g., compliance).

Other Neurodevelopmental Conditions and Conduct Disorder (CD)

ID is a neurodevelopmental disorder that overlaps with that of ADHD and autism spectrum disorder (ASD; American Psychiatric Association, 2013). A meta-analysis found a rate of ADHD in prison populations to be 26.2%, or 10-fold increase from the general population prevalence (Young, Moss, Sedgwick, Fridman, & Hodgkins, 2015). A similar rate of ADHD has been found for police detainees (Young, Goodwin, Sedgwick, & Gudjonsson, 2013). Prisoners with ADHD are at a substantially increased risk of comorbid mental health problems, primarily mood/affective and anxiety disorders, history of CD, substance abuse disorder, and personality disorder (Young, Sedgwick et al., 2015).

In a study of 90 Icelandic prisoners, ADHD was found to increase, threefold, the likelihood of a history of false confession (Gudjonsson, Sigurdsson, Einarsson, Bragason, & Newton, 2008). Similar findings have been found among large national community samples (Gudjonsson et al., 2016). The essence of these findings is that ADHD is strongly associated with false confessions in prison and community samples. The reason for this may be due to the strong comorbid relationship between ADHD and CD, particularly in an offender population (Young, Sedgwick et al., 2015), which partly drives the relationship between ADHD and false confessions (Gudjonsson et al., 2016). Drake, Gonzalez, Sigurdsson, Sigfusdottir, and Gudjonsson (2017) found that impulsivity/hyperactivity was a better predictor of false confessions than inattention, and draws a link with the recklessness and impulsivity facets of CD. Hence, the present study measured CD to adjust for any confounding influence between ADHD and false confessions, separating the inattention and impulsivity/hyperactivity dimensions.

There is absence of research into false confessions among people with ASD, although a case of a false confession to bank robbery has been reported (Perske, 2008). There is evidence that they are significantly more compliant than controls in terms of both parental and self-report (North, Russel, & Gudjonsson, 2008), leaving them more susceptible to false confessions than normal controls.

Psychiatric Symptoms

A history of mental illness has been found in fewer than 10% of false confessors (Drizin & Leo, 2004; Garrett, 2011), which is similar to the rate found among suspects detained for questioning at police stations (Gudjonsson, 2003; Gudjonsson et al., 1993; Irving, 1980; Irving & McKenzie, 1989). However, Gudjonsson et al. (1993) reported that a mental state examination revealed around 35% of police detainees had mental health problems that might interfere with their capacity to cope with the police interview. In a similar study, 29%, 24%, and 19% of suspects interrogated exhibited “severe” symptoms of anxiety, depression, and hopelessness, respectively (Sigurdsson, Gudjonsson, Einarsson, & Gudjonsson, 2006). Stress sensitivity and symptoms of depression have been shown to predict false confessions to police in community samples (Drake et al., 2017; Sigurdsson et al., 2006). Redlich, Summers, and Hoover (2010) found a 22% rate of false confessions among offenders with mental illness currently involved in the criminal justice system. Gross, Jacoby, Matheson, Montgomery, and Patil (2005) found a low rate of ID (5%) and mental illness (3%) in their sample of 340 exonerees in the United States, but among those identified as having either condition, the rate of false confession was 69%. These results suggest that ID and mental illness are salient risk factors of false confession.

Compliant Personality

Suggestibility and compliance form an intrinsic part of theories of police-induced false confessions (Gudjonsson, 2018). Compliance, as measured by the Gudjonsson Compliance Scale (GCS; Gudjonsson, 1989, 1997), has been found to predict a history of false confession in a prison sample irrespective of type (Sigurdsson & Gudjonsson, 1996), and discriminated between alleged false confessors and other court referrals (Gudjonsson, 1990). In view of this and the theoretical link between compliance and false confessions (Gudjonsson, 2003, 2018), it was included in the current study.

The present study aimed to investigate some of the most powerful predictors of false confessions, taking account of the individual predictors previously reported in the extant literature, namely, ID, ADHD, ASD, CD, psychiatric symptoms, and a compliant personality.

The hypotheses are as follows:

Hypothesis 1: There will be a significant association between the number of false confessions and mild ID, ASD, ADHD, and CD.

Hypothesis 2: There will be a significant association between the number of false confessions and psychiatric symptoms.

Hypothesis 3: There will be a significant association between the number of false confessions and compliance.

Method and Materials

Participants and Sample Selection

This study recruited 392 male inmates who were either serving a sentence or on remand at Porterfield Prison, Inverness, United Kingdom. Exclusion criteria included moderate or severe learning difficulty, a lack of fluency in the English language, and inmates being too mentally unwell to participate (as judged by prison officers). Of the 392 recruited, two participants were excluded from this secondary analysis due to essential data missing in the Diagnostic Interview for ADHD in Adults–2.0 (DIVA-2; Kooij, 2010) and a further four participants had not answered questions about false confessions. The final sample consisted of 386 male inmates who ranged in age from 18 to 50 years (M = 30.3 years, SD = 8.3 years).

Measures

Police Interrogation and Confessions Questionnaire

Participants were asked about their experiences with false confessions as follows: “Have you ever confessed during police interrogation to a criminal offense that you did not commit (i.e., you had nothing to do with the offense and are completely innocent)?” (Gudjonsson, Sigurdsson, Sigfusdottir, & Young, 2012). Replies were rated on the 5-point scale: never, once, twice, 3 to 5 times, 6 or more times, following the methodology of the original prison study (Sigurdsson, 1998; Sigurdsson & Gudjonsson, 1996a). In the current study, these count variables were coded in ordered categories.

The participants were also asked to indicate the type of offense they had falsely confessed to from the following list: “financial (theft, burglary, robbery),” “traffic violation,” “drug offense,” “sexual offense,” “criminal damage,” “violent offense,” and “other, please specify.” They then gave the reason for the false confession from the following list: “to cover up for somebody else,” “had been threatened,” “due to pressure from police,” “wanted to get away from police,” “was in alcohol/drug withdrawal,” “was taking revenge on police,” “can’t remember the reason,” and “other, please specify.”

Intelligence

IQ was estimated using the Vocabulary subtest of the Wechsler Abbreviated Scale of Intelligence (Wechsler, 1999). The Vocabulary subtest has consistently been found to be a robust predictor of general IQ and to be less susceptible to brain injury (Lezak, Howieson, Bigler, & Tranel, 2012).

Learning Disability Screening Questionnaire (LDSQ)

The LDSQ questionnaire rates an individual based on seven items that examine intellectual skills (e.g., “can the person read and write”) and functional skills (e.g., “can the person live independently”). A higher score (range = 0-7) indicates lower likelihood of ID (McKenzie, Michie, Murray, & Hales, 2012; McKenzie, Sharples, & Murray, 2015). Scores are converted to a percentage score to account for nonapplicability or nonresponses, and are then compared with a percentage cutoff, as described in the manual. The LDSQ has been reported to have good convergent validity when compared with Wechsler IQ scores (McKenzie et al., 2015), with more than 80% specificity and sensitivity, and has been validated for use in forensic settings, where it is also found to have acceptable psychometric properties (McKenzie et al., 2012).

ADHD diagnosis

All participants were interviewed using the DIVA-2 (Kooij, 2010), which is a validated semistructured clinical interview of ADHD in adults. ADHD consists of two underlying domains (a) inattention and (b) hyperactivity/impulsivity. The interview questions address participants’ current and childhood (i.e., before age 12) presentation of symptoms, and inmates were classified according to Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association, 2013) criteria, which requires a person suffers with associated impairments.

The Autism Quotient (AQ)

The AQ is a self-reported screening instrument of ASD with good reliability and validity (Baron-Cohen, Wheelwright, Skinner, Martin, & Clubley, 2001; Stevenson & Hart, 2017). One point is awarded for each response that indicates behaviors within the autistic spectrum. Total scores range from 0 to 50 with higher scores reflecting the presence of a greater number of autistic symptoms. The screening cutoff value of 26 (Hoekstra, Bartels, Cath, & Boomsma, 2008) was applied for this research, and inmates who scored above this cutoff were classified as likely having a general ASD.

CD screening

CD was screened using the CD scale of the Barkley Adult ADHD Rating Scale–IV (Barkley, 2011), which corresponds with DSM-5 criteria, with endorsement of three or more criteria indicating likelihood of the disorder.

Brief Symptom Inventory (BSI®)

The BSI is a self-report scale that measures psychological distress and psychiatric disorder (Derogatis, 1993). The BSI has nine subscales (i.e., Somatization, Obsessive–Compulsive, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychoticism), and three composite measures: global severity index (i.e., the mean of all the subscale scores), positive symptoms total (i.e., the number of items endorsed at a higher level than 0), and positive symptom distress (i.e., the sum of items values divided by positive symptoms total score). Reliability and validity of the BSI have been readily documented (Derogatis, 1993).

GCS

The GCS (Gudjonsson, 1989, 1997) is a well-established 20-item self-report measure of compliance and is comprised of two main factors: an eagerness to please people in authority and avoidance of conflict and confrontation. It correlates positively with symptoms of anxiety, low self-esteem, and paranoia (Gudjonsson, Sigurdsson, Brynjolfsdottior, & Hreinsdottir, 2002).

Procedures

The all-male sample was recruited by opportunity sampling from Porterfield (United Kingdom) over a period of 18 months. Those who indicated interest attended an appointment with the researcher when they were given detailed written information about the study. Following consent, an appointment was made to administer the battery of measures. This took approximately 4 hr (usually split across two or three sessions). Of the 386 participants, 79 (20.5%) required help with reading the questionnaires. The two researchers received comprehensive training to administer these measures; and, for the DIVA-2 training, they were required to watch and score video recordings of patients to ensure reliability.

The procedure of the original study was reviewed and ethically approved by the Scottish Prison Service Research Access and Ethics Committee (No. 7/13/10/10). Further details of the study procedures have been published elsewhere (Young et al., 2016).

Analytical Strategy

Frequencies were reported for all categorical variables, and means with their standard deviations for continuous descriptive variables. To establish independence in the proportions of the observations of all binary and categorical variables, we used chi-square tests and measure effect sizes by the use of odds ratio (OR) and Cramer’s V.

Taking into account the proportion of zero responses, and their overall distribution, we treated the response variable number of false confessions as count, and fitted negative binomial regression (NBR) models. A Poisson distribution is appropriate in accounting for observed heterogeneity when using count data but is not when there is overdispersion, as observed in false confessions (Long & Freese, 2006). NBR models were fitted to examine the association of neurodevelopmental exposure variables, ADHD severity/dimensional measures, and BSI symptom with number of false confessions. We further analyzed the relationship between compliance and number of false confessions, and the potentially explanatory role of either hyperactivity/impulsivity or CD on associations between compliance and number of false confessions, using NBR models.

Model beta coefficients were exponentiated, with incidence rate ratios (IRRs) as the measure of magnitude in NBRs. A significance level of α = .05 was adopted throughout. All analyses were performed using Stata version 13 (StataCorp., 2013).

Results

Table 1 shows the frequencies of different developmental risk factors for false confessions. Among those who had made a false confession ever (129, 33.4%), 54 (41.9%) confessed once, 37 (28.7%) confessed twice, 28 (21.7%) did it 3 to 5 times, and 10 (7.8%) had falsely confessed 6 times or more.

Table 1.

Distribution of Any False Confessions Among the Four Diagnostic Groups.

Diagnostic group	All inmatesn = 386n (%)	No false confession(n = 257, 66.6%)	False confession(n = 129, 33.4%)	χ² (df)	p
ID	35 (9.1)	21 (8.2)	14 (10.9)	0.7 (1)	.393
ASD	33 (8.5)	17 (6.6)	16 (12.4)	3.7 (1)	.055
ADHD	95 (24.6)	53 (20.6)	42 (32.6)	6.6 (1)	.010
CD	153 (39.6%)	86 (33.5)	67 (51.9)	12.3(1)	.000

Note. ID = intellectual disability; ASD = autism spectrum disorder; CD = conduct disorder.

The offenses falsely confessed to were property (financial) offenses (37, 28.7%), drug offenses (20, 15.5%), violent offenses (18, 14.0%), criminal damage (17, 13.2%), serious traffic violations (10, 7.8%), sex offenses (3, 2.3%), and “other” (24, 18.6%). The “other” category included breach of the peace, car theft, domestic violence, house breaking, and vandalism.

Of the 127 participants who reported making a false confession and provided the reason (two did not complete the relevant section), 79 (62.2%) reported the main reason was to “cover up for somebody else.” This was followed by “pressure from police” (18, 14.2%), “to get away from police” (5, 3.9%), “alcohol/drug withdrawal” (2, 1.6%), and “had been threatened” (1, <1%). Three could not remember the reason (2.4%) and 19 (15.0%) said it was for some “other” reason. All 19 described a motive, which included a perception of a reduced sentence if confessed (n = 6); wanted to get out of the police station more quickly (n = 3); was interviewed when intoxicated (n = 2); to get support/wanted to go to Borstal to get out of a care home (n = 2); told to confess to police or lawyer (n = 2); needed to be punished, because felt guilty for another crime he had truly committed (n = 1); did not want to “grass” on the real perpetrator (n = 1); “panicked” (n = 1); and there was “false evidence” against him (n = 1).

Neurodevelopmental Conditions and CD

Table 1 shows that a history of false confessions was significantly more common among those who met diagnostic criteria for ADHD (32.6%) than those who did not (20.6%), p = .01, OR = 1.86, confidence interval (CI) = [1.15, 2.99], Cramer’s V = 0.131. Inmates who met screening criteria for CD (51.9%) were significantly more likely to report a history of false confessions than those who did not (33.5%; p < .001, OR = 2.15, CI = [1.40, 3.31], Cramer’s V = 0.178). Both represent a medium effect size.

Those screening positive for ID and ASD did not have significantly higher rates of inmates who falsely confessed, although there was a trend for ASD (p = .055, OR = 2.00, CI = [0.97, 4.10], Cramer’s V = 0.098).

There were no significant mean differences in estimated IQ among inmates who falsely confessed from those who did not. Also, there was no significant difference in age between the two groups. However, 46 (25.6%) of the false confessors, in contrast to 33 (17.9%) of the other prisoners, required help with reading the questionnaires. The difference was not significant, but showed a trend that should be investigated in future studies (χ² = 3.118, df = 1, p = .083).

We aimed to identify the contribution of symptom domains of inattention and hyperactivity/impulsivity on false confessions (Table 2). These results reveal that hyperactivity/impulsivity but not inattentiveness drove the association between ADHD symptoms and number of false confessions, after adjusting for coexisting CD.

Table 2.

Association Between Severity of ADHD Symptoms and Number of False Confessions.

	Univariate	Adjusted^a	Adjusted^b
Count	IRR [90% CI]	aIRR [90%CI]	aIRR [90% CI]
Inattention	1.07 [1.01, 1.14]*	1.04 [0.98, 1.12]	0.99 [0.91, 1.07]
Hyperactivity/impulsivity	1.12 [1.05, 1.21]**	1.10 [1.02, 1.18]*	1.11 [1.01, 1.21]*
Total symptoms	1.06 [1.02, 1.10]**	1.04 [1.00, 1.08]*	—

Note. IRR = incidence rate ratio; CI = confidence interval; aIRR = adjusted incidence rate ratio; CD = conduct disorder.

Adjusted for CD.

Adjusted for CD and the inattention ADHD symptom dimension.

p < .01. **p < .001.

Psychiatric Symptoms

Table 3 presents models examining the association between BSI psychiatric symptom domains and number of false confessions. Somatization, obsessive–compulsive, anxiety, hostility, paranoid, and psychoticism were all related to false confessions. The largest mean difference between those who falsely confessed or not was observed on the global severity index (p < .01).

Table 3.

Association Between Number of False Confessions and Psychiatric Symptom Domains.

BSI	False confessions
BSI	Incidence rate ratio	95% confidence interval	p
Somatization	1.02	[1.00, 1.03]	.014
Obsessive–compulsive	1.02	[1.00, 1.03]	.020
Interpersonal–sensitivity	1.01	[0.99, 1.02]	.480
Depression	1.01	[0.99, 1.03]	.152
Anxiety	1.02	[1.00, 1.03]	.012
Hostility	1.02	[1.00, 1.03]	.008
Phobic anxiety	1.01	[0.99, 1.02]	.348
Paranoid	1.02	[1.00, 1.04]	.013
Psychoticism	1.02	[1.00, 1.03]	.035
Global Severity Index	1.02	[1.01, 1.04]	.003

Note. BSI = Brief Symptom Inventory.

p < .01. **p < .001.

Compliant Personality

Compliance was significantly associated with the number of false confessions (IRR = 1.07, p < .05, 90% CI = [1.02, 1.13]). Further analyses did not indicate a mediating or moderating role of compliance in the association between hyperactivity/impulsivity or CD and number of false confessions.

Discussion

The rate of reported false confessions (33.4%) in this study is much higher than those reported in two similar Icelandic studies, where the rates of false confession were 12.2% and 24.4%, respectively (Gudjonsson et al., 2008; Sigurdsson & Gudjonsson, 1996a). The two Icelandic studies were conducted over 10 years apart, using the same methodology, and the rate of false confession had doubled over that period. Gudjonsson et al. (2008) explained the differences over time in terms of false confessions, either being on the increase in Iceland or that participants are now more aware of it and more willing to report it. The methodology used in the Scottish study was modeled on the Icelandic studies and the results are, therefore, comparable. The rate of false confession and its causes and motivation may vary across countries and jurisdictions (Gudjonsson, 2018) but it is also possible that false confessions, or a willingness to report them, are also increasing the phenomenon.

The majority of the false confession sample reported giving a false confession on more than one occasion, suggesting a behavioral pattern associated with their lifestyle. NBR models were fitted to examine the association of neurodevelopmental exposure variables, ADHD severity/dimensional measures, and BSI symptom with number of false confessions. This gave a more sophisticated analysis of the false confession data than merely relying of “false confessions” versus “no false confessions.”

The participants were serving prisoners, who had in 62.2% of cases reportedly given a false confession to protect the real perpetrator rather than being police coerced. These types of “voluntary” false confession rarely feature in cases of miscarriage of justice, because they are typically undisputed and are “hidden” within the criminal justice system (Gudjonsson, 2018). They are best construed as forming a part of a delinquent/criminal lifestyle, driven significantly by CD and ADHD symptoms rather than the inability of the suspect to cope with the police interview (Gudjonsson et al., 2016).

CD and ADHD were the most powerful predictors of false confessions in the study (ORs = 2.15 and 1.86, respectively). CD mediated some of the ADHD effects once this had been adjusted for in the analyses. This supports previous findings among prisoners and community samples, which show false confessions to be associated with peer delinquency (Gudjonsson, Sigurdsson, Asgeirsdottir, & Sigfusdottir, 2006), CD (Gudjonsson et al., 2016), antisocial personality traits (Gudjonsson, Sigurdsson, & Einarsson, 2004), and involvement in delinquency/criminal activity (Gudjonsson, Sigurdsson, & Sigfusdottir, 2009; Redlich, Kulish, & Steadman, 2011; Sigurdsson & Gudjonsson, 2001). These factors increase the likelihood of police involvement and interrogation, leading to a higher risk of false confession (Gudjonsson et al., 2016).

The hyperactivity/impulsivity ADHD symptom dimension was a better predictor of false confessions than the inattention symptom dimension, supporting the epidemiological/community findings of Drake et al. (2017), and remained significant after adjusting for the confounding effects of CD. These are robust effects and the key components of ADHD associated with false confessions, as Drake et al. (2017) suggest, may be recklessness and lack of self-control (i.e., impulsivity).

Compliance was significantly associated with the reported rate of false confession, but it did not mediate or moderate the relationship of ADHD and CD with false confessions. This may, in part, be due to the nature of the false confessions reported in the current study. The majority of reasons were to cover up an offense for another person rather than a consequence of police pressure or threats from the real offender (i.e., they were voluntary), which is consistent with the findings of Sigurdsson (1998), where only two (7.1%) of the “protecting somebody else” false confessions were the results of threats by the real offender.

Covering up an offense as a favor to another person is not associated with compliance, in contrast to pressure from the guilty person to take the blame (Gudjonsson, Sigurdsson, & Einarsson, 2007). The relationship between compliance and false confessions is most relevant to coerced confessions (Gudjonsson, 2018).

We found an association between number of false confessions and psychiatric symptoms, but the effect sizes were small. The largest mean difference between those who falsely confessed and not was observed on the global severity index, which suggests that poor general mental health is a risk factor to false confessions. A limitation is that the BSI measures current symptoms, which may or may not have been present to the same extent when the false confessions were made. In the present study, the false confessions were not dated and any mental health issues, including psychiatric symptoms, at the time of making the false confession could not be ascertained. This is not a problem in relation to ID and ASD, which are persistent conditions. ADHD symptoms do remit with age, but about 65% of those diagnosed in childhood who were symptomatic at the age of 25 retain persistent symptoms associated with significant functional impairments (Faraone, Biederman, & Mick, 2006; Young & Gudjonsson, 2008).

We found that neither ID nor prorated verbal IQ predicted false confessions. The rate of ID for the entire sample was 9.1%, which is consistent with the rate found among prisoners (Hayes et al., 2007) and suspects detained for interviews at police stations (Gudjonsson et al., 1993; Young et al., 2013). The findings corroborate those of Sigurdsson and Gudjonsson (1996a) that IQ is not a significant predictor of false confession history among prisoners. The findings suggest that false confessors are no more intellectually disadvantaged than other police detainees or prisoners, but both groups have a disproportionate number of people with ID, requiring the services of an “appropriate adult” during police questioning (Medford, Gudjonsson, & Pearse, 2003).

There is an important caveat to consider, however. Studies involving interviews and completion of psychometric tests typically exclude cases of moderate and severe ID for ethical and pragmatic reasons, which leaves the possibility that those cases may involve disproportionate number of false confessions. Support for this is found in cases of 340 exonerations in the United States between 1989 and 2003 (Gross et al., 2005).

Prisoners with ASD did not have a significantly higher rate of false confession than other prisoners, although ASD was close to significance (p = .055). Therefore, a measure of ASD should be included in future research into false confessions.

There are some limitations that need to be considered with regard to the findings. Prisoners with moderate and severe ID were excluded from the study and these may be the persons most vulnerable to making false confessions. Gudjonsson (2018) presents evidence to show that ID impairs the capacity of suspects to understand their legal rights, ability to exercise their rights, and to make informed decisions. The relationship between ID and giving a false confession is complex, reflecting motivational factors; coexisting vulnerabilities such as suggestibility, compliance, and acquiescence; how they are interviewed; and the support they receive while at the police station (e.g., access to a lawyer and an appropriate adult).

A further limitation is that we do not know whether the current clinical characteristics are related to the propensity to confess falsely specifically rather than to confessing generally (i.e., either truly or falsely). The current study only focused on the history of false confessions, and no comparison could be made with those who had given true confessions. The available evidence is that true confessions are primarily related to the perception of proof and internal pressure to confess during interrogation, whereas false confessions are more related to the perception of external pressure and short-term gains (Redlich et al., 2011; Sigurdsson & Gudjonsson, 2006b). False confessors also tend to be interrogated more often and for longer than true confessors (Gudjonsson, 2003; Redlich et al., 2011).

In terms of their personality, true confessors and true deniers have the most normal personality pattern, with false confessors and false deniers exhibiting more antisocial personality traits (Gudjonsson, Sigurdsson, Bragason, et al., 2004).

A confounding variable not investigated in the current study is that the risk of false confession is likely to be related to the perceived severity of punishment, making it generally easier to confess falsely to minor than serious offenses. However, there have been some reported serial false confessors to serious crimes, such as murder (Gudjonsson, 2003).

Research has shown that false confessions are typically caused by a combination of situational and personal factors and the more serious the alleged offense, the longer and more intense the interrogation tends to be (Gudjonsson, 2018). Gudjonsson (2018) reviews 17 different risk factors to false confession; typically more than one applies in a given case and two or more factors may operate individually or in combination. One of the more recently recognized risk factors is the “mind-set” of the suspect prior to and during interrogation (e.g., naivety about the criminal justice process and the primary focus on the immediate rather than long-term consequences of the confession).

The most predominant reason given for the false confession was to “cover up for somebody else,” which was reported in 62.2% of cases. This differs from most published cases of miscarriage of justice, which typically involve police-induced false confessions, and where ID probably plays a more prominent role in terms of impact (Drizin & Leo, 2004; Garrett, 2011; Gross et al., 2005; Gudjonsson & MacKeith, 1994; Perske, 2008). The history of false confession was based on self-report, it was not independently corroborated and they did not necessarily occur in close proximity to the completion of the measure.

The strengths of the study are the large sample size and the inclusion of the key risk factors: ID, ADHD, ASD, CD, current psychiatric symptoms, and compliance. This makes the study unique. The retrospective account of the false confessions may have reduced the detail and accuracy of the recall of the event and surrounding circumstances. In addition, no measures were available of the mental state of the prisoners at the time they gave the false confessions, which typically forms the part of the evaluation in cases of miscarriage of justice involving disputed confessions (Gudjonsson, 2003).

In conclusion, the study highlighted the important role of CD and ADHD in giving a false confession but the risk is likely mediated by context, that is, the reason for giving a false confession. Nevertheless, we conducted a secondary analysis and, although we had the opportunity to include several risk factors, there may be others that were not included in this analysis.

Footnotes

Authors’ Note

Gisli Hannes Gudjonsson is currently affiliated with King’s College London, England.

Susan Young is currently affiliated with Psychology Services Limited, England, UK.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: SY has received honoraria for consultancy, travel, educational talks and/or research from Janssen, Eli Lilly, HB Pharma, and/or Shire. GHG and RG have no conflicts of interest.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was supported by Shire Pharmaceutical Development Limited through a restricted grant. Shire had no role in the design and conduct of the study (collection, management, analysis, and interpretation of the data) or on the preparation, review, or approval of the manuscript, and the decision to submit the manuscript for publication. None of the authors received funds for their involvement in this manuscript.

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Gisli Hannes Gudjonsson, CBE, PhD is a registered clinical and forensic psychologist. He is a professor of psychology at Reykjavik University and an emeritus professor of forensic psychology at the Institute of Psychiatry, Psychology and Neuroscience, King’s College London.

Rafael A. González has a PhD in clinical psychology. He works in a specialist ADHD Child and Adolescent Mental Health Service in East London. His research interests include ADHD and its comorbidity with substance use and dependence.

Susan Young, BSc, DClinPsy, PhD, CSi, AFBPS is a visiting professor at Reykjavik University. She is director of Psychology Services Limited (). She has developed and published numerous clinical assessment tools and psychological programmed interventions; her work has been published in over 20 languages.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Barkley

R. A.

(2011). Barkley Adult ADHD Rating Scale-IV (BAARS-IV). New York, NY: Guilford Press.

Baron-Cohen

Wheelwright

Skinner

Martin

Clubley

(2001). The autism-spectrum quotient (AQ): Evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders, 31, 5-17.

Derogatis

L. R.

(1993). BSI®. Brief Symptom Inventory: Administration, scoring, and procedures manual. San Antonio, TX: Pearson.

Drake

K. E.

Gonzalez

R. A.

Sigurdsson

J. F.

Sigfusdottir

I. D.

Gudjonsson

G. H.

(2017). A national study into temperament as a critical susceptibility factor for reported false confessions amongst adolescents. Personality and Individual Differences, 111, 220-226.

Drizin

S. A.

Leo

R. A.

(2004). The problem of false confessions in the post-DNA world. North Carolina Law Review, 82, 891-1007.

Faraone

S. V.

Biederman

Mick

(2006). The age-dependent decline of attention deficit hyperactivity disorder: A meta-analysis of follow-up studies. Psychological Medicine, 36, 159-165.

Garrett

B. L.

(2011). Convicting the innocent: Where criminal convictions go wrong. London, England: Harvard University Press.

Gross

S. R.

Jacoby

Matheson

D. J.

Montgomery

Patil

(2005). Exonerations in the United States, 1989 through 2003. Journal of Criminal Law and Criminology, 95, 523-553.

10.

Gudjonsson

G. H.

(1989). Compliance in an interrogation situation: A new scale. Personality and Individual Differences, 10, 535-540.

11.

Gudjonsson

G. H.

(1990). One hundred alleged false confession cases: Some normative data. British Journal of Clinical Psychology, 29, 249-250.

12.

Gudjonsson

G. H.

(1997). The Gudjonsson suggestibility scales manual. Hove, UK: Psychology Press.

13.

Gudjonsson

G. H.

(2003). The psychology of interrogations and confessions: A handbook. Chichester, UK: John Wiley.

14.

Gudjonsson

G. H.

(2010). Psychological vulnerabilities during police interviews: Why are they important? Legal and Criminological Psychology, 15, 161-175.

15.

Gudjonsson

G. H.

(2018). The psychology of false confessions: Forty years of science and practice. Chichester, UK: Wiley Blackwell.

16.

Gudjonsson

G. H.

Clare

Rutter

Pearse

(1993). Persons at risk during interviews in police custody: The identification of vulnerabilities. London, England: Royal Commission on Criminal Justice, Her Majesty’s Stationery Office.

17.

Gudjonsson

G. H.

MacKeith

(1994). Learning disability and the Police and Criminal Evidence Act 1984.Protection during investigative interviewing: A video-recorded false confession to double murder. Journal of Forensic Psychiatry, 5, 35-49.

18.

Gudjonsson

G. H.

Sigurdsson

J. F.

Asgeirsdottir

B. B.

Sigfusdottir

I. D.

(2006). Custodial interrogation, false confession and individual differences. A national study among Icelandic youth. Personality and Individual Differences, 41, 49-59.

19.

Gudjonsson

G. H.

Sigurdsson

J. F.

Bragason

O. O.

Einarsson

Valdimarsdottir

E. B.

(2004). Confessions and denials and the relationship with personality. Legal and Criminological Psychology, 9, 121-133.

20.

Gudjonsson

G. H.

Sigurdsson

J. F.

Brynjolfsdottior

Hreinsdottir

(2002). The relationship of compliance with anxiety, self-esteem, paranoid thinking and anger. Psychology, Crime & Law, 8, 145-153.

21.

Gudjonsson

G. H.

Sigurdsson

J. F.

Einarsson

(2004). The role of personality in relation to confessions and denials. Psychology, Crime & Law, 10, 125-135.

22.

Gudjonsson

G. H.

Sigurdsson

J. F.

Einarsson

(2007). Taking blame for antisocial acts and its relationship with personality. Personality and Individual Differences, 43, 3-13.

23.

Gudjonsson

G. H.

Sigurdsson

J. F.

Einarsson

Bragason

O. O.

Newton

A. K.

(2008). Interrogative suggestibility, compliance and false confessions among prisoners and their relationship with attention deficit hyperactivity disorder (ADHD) symptoms. Psychological Medicine, 38, 1037-1044.

24.

Gudjonsson

G. H.

Sigurdsson

J. F.

Sigfusdottir

I. D.

(2009). Interrogations and false confessions among adolescents in seven countries in Europe. What background and psychological factors best discriminate between false confessors and non-false confessors? Psychology, Crime & Law, 15, 711-728.

25.

Gudjonsson

G. H.

Sigurdsson

J. F.

Sigfusdottir

I. D.

Asgeirsdottir

B. B.

González

Young

(2016). A national epidemiological study investigating risk factors for police interrogation and false confession among juveniles and young persons. Social Psychiatry and Psychiatric Epidemiology, 51, 359-367. doi:10.1007/s00127-015-1145-8

26.

Gudjonsson

G. H.

Sigurdsson

J. F.

Sigfusdottir

I. D.

Young

(2012). False confessions to police and their relationship with conduct disorder, ADHD, and life adversity. Personality and Individual Differences, 52, 696-701.

27.

Hayes

Shackell

Mottram Lancaster

(2007). The prevalence of intellectual disability in a major UK prison. British Journal of Learning Disabilities, 35, 162-167.

28.

Hoekstra

R. A.

Bartels

Cath

D. C.

Boomsma

D. I.

(2008). Factor structure, reliability and criterion validity of the Autism-Spectrum Quotient (AQ): A study in Dutch population and patient groups. Journal of Autism and Developmental Disorders, 38, 1555-1566.

29.

Irving

B. L.

(1980). Police interrogation: A case study of current practice (Research Study No. 2). London, England: Her Majesty’s Stationery Office.

30.

Irving

B. L.

McKenzie

I. K.

(1989). Police interrogation: The effects of the Police and Criminal Evidence Act 1984. London, England: The Police Foundation.

31.

Kooij

J. J. S.

(2010). Diagnostic Interview for ADHD in Adults 2.0 (DIVA 2.0). In Kooij

J. J. S.

(Ed.), Adult ADHD: Diagnostic assessment and treatment (pp. 33-99). Amsterdam, The Netherlands: Pearson Assessment and Information BV.

32.

Lezak

M. D.

Howieson

D. B.

Bigler

E. D.

Tranel

(2012). Neuropsychological assessment (5th ed.). New York, NY: Oxford University Press.

33.

Long

J. S.

Freese

(2006). Regression models for categorical dependent variables using Stata, Second edition. College Station, TX: Stata Press.

34.

McKenzie

Michie

Murray

Hales

(2012). Screening for offenders with an intellectual disability: The validity of the Learning Disability Screening Questionnaire. Research in Developmental Disabilities, 33, 791-795. doi:10.1016/j.ridd.2011.12.006

35.

McKenzie

Sharples

Murray

A. L.

(2015). Validating the Learning Disability Screening Questionnaire against the Weschler Adult Intelligence Scale, Fourth Edition. Intellectual and Developmental Disabilities, 53, 301-307. doi:10.1352/1934-9556-53.4.301

36.

Medford

Gudjonsson

G. H.

Pearse

(2003). The efficacy of the appropriate adult safeguard during police interviewing. Legal and Criminological Psychology, 8, 253-266.

37.

North

A. S.

Russel

A. J.

Gudjonsson

G. H.

(2008). High functioning autism spectrum disorders: An investigation of psychological vulnerabilities during interrogative interview. Journal of Forensic Psychiatry & Psychology, 19, 323-334.

38.

Perske

(2008). False confessions from 53 persons with intellectual disabilities: The list keeps growing. Intellectual and Developmental Disabilities, 46, 468-479.

39.

Redlich

A. D.

Kulish

Steadman

H. J.

(2011). Comparing true and false confessions among persons with serious mental illness. Psychology, Public Policy, and Law, 17, 394-418.

40.

Redlich

A. D.

Summers

Hoover

(2010). Self-reported false confessions and false pleas among offenders with mental illness. Law and Human Behavior, 34, 79-90.

41.

Schatz

S. J.

(2018). Interrogated with intellectual disabilities: The risks of false confessions. Stanford Law Review, 70, 643-690.

42.

Sigurdsson

J. F.

(1998). Alleged false confessions among Icelandic offenders: An examination of some psychological, criminological and substance use factors that are associated with the reported false confessions (Doctoral thesis). King’s College London.

43.

Sigurdsson

J. F.

Gudjonsson

G. H.

(1996a). Psychological characteristics of “false confessors.” A study among Icelandic prison inmates and juvenile offenders. Personality and Individual Differences, 20, 321-329.

44.

Sigurdsson

J. F.

Gudjonsson

G. H.

(1996b). The relationship between types of claimed false confession made and the reasons why suspects confess to the police according to the Gudjonsson Confession Questionnaire (GCQ). Journal of Criminological and Legal Psychology, 1, 259-269.

45.

Sigurdsson

J. F.

Gudjonsson

G. H.

(2001). False confessions: The relative importance of psychological, criminological and substance abuse variables. Psychology, Crime & Law, 7, 275-289.

46.

Sigurdsson

J. F.

Gudjonsson

G. H.

Einarsson

Gudjonsson

(2006). Differences in personality and mental state between suspects and witnesses immediately after being interviewed by the police. Psychology, Crime & Law, 12, 619-628.

47.

StataCorp. (2013). Stata statistical software: Release 13. College Station, TX: Author.

48.

Stevenson

J. L.

Hart

K. R.

(2017). Psychometric properties of the Autism-Spectrum Quotient for assessing low and high levels of autistic traits in college students. Journal of Autism and Developmental Disorders, 47, 1838-1853.

49.

Viljoen

J. L.

Klaver

Roesch

(2005). Legal decisions of preadolescent and adolescent defendants: Predictors of confessions, pleas, communication with attorneys, and appeals. Law and Human Behavior, 29, 253-277.

50.

Wechsler

(1999). Wechsler Abbreviated Scale of Intelligence. New York, NY: The Psychological Corporation.

51.

Young

Gonzalez

R. A.

Mutch

Mallet-Lambert

O’Rourke

Hickey

. . . Gudjonsson

G. H.

(2016). Diagnostic accuracy of a brief screening tool for attention deficit hyperactivity disorder in UK prison inmates. Psychological Medicine, 46, 1449-1458. doi:10.1017/S0033291716000039

52.

Young

Goodwin

E. J.

Sedgwick

Gudjonsson

G. H.

(2013). The effectiveness of police custody assessments in identifying suspects with intellectual disabilities and attention deficit hyperactivity disorder. BMC Medicine, 11, Article 248.

53.

Young

Gudjonsson

G. H.

(2008). Growing out of ADHD: The relationship between functioning and symptoms. Journal of Attention Disorders, 12, 162-169.

54.

Young

Moss

Sedgwick

Fridman

Hodgkins

(2015). A meta-analysis of the prevalence of attention deficit hyperactivity disorder in incarcerated populations. Psychological Medicine, 45, 247-258.

55.

Young

Sedgwick

Fridman

Gudjonsson

Hodgkins

Lantigua

González

(2015). Co-morbid psychiatric disorders among incarcerated ADHD populations: A meta-analysis. Psychological Medicine, 45, 2499-2492.