Generalizability of a two-factor measure of young children's suggestibility in Norway and the USA

Abstract

We discuss the development of the Book Suggestibility Scale for Children (BSSC), a two-factor suggestibility measure developed as an alternative to the Video Suggestibility Scale for Children (VSSC; Scullin & Ceci, Personality and Individual Differences, 30, 843–856, 2001). In study 1, both the BSSC and the VSSC were administered to 60 American preschool children. Children responded affirmatively to inaccurate leading questions (Yield) and changed their answers in response to negative feedback from interviewers (Shift) in a similar manner on both scales. In study 2 we found that the BSSC was useful for studies of suggestibility in contexts other than America because 62 Norwegian preschool children responded to questions in a similar manner to an age-range and vocabulary matched group of 31 American children. Results support a two-factor model of suggestibility, although Yield appears to measure suggestibility across scales and situations more consistently than does Shift. American age-matched samples had significantly higher Shift and Yield 2 scores than the Norwegian sample.     

Memory conformity and suggestibility

Abstract

The present study examined whether the degree to which participants engage in memory conformity, which occurs when a person alters their memory report of an event to be consistent with another person, can be predicted by their levels of interrogative suggestibility (IS), which is the degree to which people are susceptible to altering their memory reports during questioning. Memory conformity was introduced by having participant and confederate pairs study words and then complete a social recognition test where they took turns to make judgements to the same items. When the participants responded after the confederate, they tended to conform to confederate's judgements regardless of whether the confederate had made a correct or incorrect response. IS was measured using the Gudjonsson Suggestibility Scale 2. This scale allows separate measures of Yield, which is a measure of how susceptible people are to altering their memory reports of events as a result of leading questions, and Shift, which is a measure of how susceptible people are to changing responses to questions when placed under pressure to do so. Only Yield was a significant predictor of memory conformity.     

Interacting influences on interrogative suggestibility

Purpose. Research using the Gudjonsson Suggestibility Scales (GSS) has found interrogative suggestibility (IS) to vary as a function of the overall demeanour of the interviewer, warnings about the presence of misleading information, and the self‐esteem of the interviewee, as outlined by Gudjonsson and Clark (1986). The present study attempted to assess how these factors interact. Method. The study had a three‐factor between‐participants design: interviewer demeanour × instructional manipulation × self‐esteem. One hundred and twenty undergraduates took part in the study. Results. Participants reporting lower self‐esteem scored higher on the GSS ‘Shift’ measure than participants reporting higher levels of self‐esteem. Participants faced with a ‘Friendly’ interviewer scored lower on the GSS Yield 1 and Total Suggestibility measures than did those participants faced with a more ‘Abrupt’ interviewer. Participants warned about the presence of misleading information scored lower on Yield 1 and Total Suggestibility. A potentially key finding was that participants who received a warning demonstrated an increased number of Shifts in the Friendly condition compared with those who were not warned. In the Abrupt condition this pattern was reversed. Conclusion. The results supported studies showing that all three variables tested affect levels of IS but further suggested that optimal interviewer support for interviewees' discrepancy detection may be provided either by a relaxed interviewer manner or by warnings alone, but not by both.     

Negative emotional states,life adversity,and interrogative suggestibility

Purpose. Interrogative suggestibility has been shown to vary according to cognitive and personality factors and if reliably measured may predict performance in real forensic interviews. It is therefore of both theoretical and practical interest to identify which psychological factors are most closely related to suggestible responding. This study examines the extent to which individual differences in negative emotional states predict performance on a measure of interrogative suggestibility and also tests the assumption that self‐reports of negative life events are associated with suggestibility. Method. A non‐clinical sample (N= 80) of participants was administered the brief form of the Depression Anxiety Stress Scales (DASS‐21), the Life Experiences Survey (LES), and the Gudjonsson Suggestibility Scale (GSS‐1). Results. Negative emotional states were found to correlate positively, although moderately, with all of the suggestibility measures. Multiple regression analyses found significant predictive models emerged for Yield 1, Yield 2, and Total Suggestibility. Each of these predicted a small proportion of the variance. Negative life event impact ratings were not associated with interrogative suggestibility. Conclusions. The findings suggest that brief self‐report measures of negative emotional states are limited as predictors of interrogative suggestibility. The results also call into question the predictive utility of traditional checklist measures of life adversity for forensic purposes.     

Suggestibility and children with mild learning disabilities: The use of the cognitive interview

Abstract

The aims of this paper were firstly to identify any differences in the level of suggestibility between 20 7–9-year-old children with mild learning disabilities and 20 children with average academic ability using the Gudjonsson Suggestibility Scale 2 (GSS2) and, secondly, to note the impact of the cognitive interview on the response patterns of children with mild learning disabilities.

On the GSS2, average academic ability children recalled significantly more correct details than children with mild learning disabilities. There was no significant difference between the two groups of children on distortions, fabrications, total confabulations or on any of the four measures of suggestibility.

In the second part of the study, 38 children with mild learning disabilities watched a filmed event and then were interviewed using either a standard or a modified version of the cognitive interview. Following this, all participants were asked specific questions, some of which incorporated misleading information. One week later each participant was re-questioned. This time critical questions were included about the truth of the presuppositions introduced in the initial interview session. It was found that the cognitive interview elicited significantly more correct and incorrect details than a standard interview with no significant difference in fabrications. Use of the cognitive interview did not significantly affect susceptibility to subsequent misleading suggestions.     

Age trends in eye-witness suggestibility and compliance

Abstract

This paper addresses two issues: whether there is a developmental trend in suggestibility to misleading post-event information, and whether suggestibility can be reduced by use of part of the cognitive interview. Twenty participants from each of three age groups (four-to-five year olds, eight-to-nine year olds and adults) watched a filmed event, and half of the participants in each age group were subsequently asked to recall everything they had seen using a method derived from the cognitive interview procedure. Following this, all participants were asked questions about the filmed event, some of which incorporated misleading information. Twenty four hours later the witnesses were interviewed again, this time critical questions were included about the truth of the presuppositions introduced in the initial questionnaire. It was found that although the eight-to-nine year olds were more suggestible than adults, the apparent greater suggestibility of very young children (four-to-five years) could potentially be explained in terms of heightened compliance to the perceived demands of the interviewer. The 'be complete' part of the cognitive interview only produced an improvement in performance for the eight-to-nine year olds.     

Differences in suggestibility between 5–6 and 10–11 year olds: The relationship with self confidence

Abstract

Differences in suggestibility and recall between 5–6 and 10–11 year olds were investigated. It was hypothesized that younger children would be more suggestible than older children and that differences in self confidence between older and younger children would influence these differences. It was also predicted that older children would recall more information than younger children, and that this age difference would be less strongly influenced by self confidence. Forty-one 5 and 6 and fifty-six 10 and 11 year olds were interviewed about a video they had witnessed. Several factual and misleading questions were asked, and the percentage of correct answers to the factual questions (to measure recall) and the percentage of incorrect answer to the misleading questions (to measure suggestibility) were calculated. The level of self confidence of the children was measured with six items of the Behavioural Academic Self Esteem Scale (BASE), reflecting self confidence. The outcomes supported the hypotheses: Younger children were more suggestible than older children and this difference disappeared when controlled for self confidence. Older children gave mere information about the event than younger children, and these age differences were, to much less extent, influenced by their self confidence.     

Interrogative suggestibility: Was it just compliance or a genuine false memory?

Purpose. Interrogative suggestibility, as measured by the Gudjonsson Suggestibility Scales (GSS), is an independent form of suggestibility arising in the forensic/legal context. So far, an unresolved issue that may have different implications when measuring suggestibility is to what extent the scales measure internalization of suggested materials or just compliance with the interrogator. Methods . Internalization of suggested materials and compliance were here measured using a source identification task. In Experiment 1, participants were administered the GSS2 and immediately afterwards asked to perform the source identification task on the items presented in the scale. In Experiment 2, half of the participants were administered the source identification task immediately and half after a 24‐hr delay. Results . In both experiments a higher proportion of compliant responses were found. Participants internalized more suggested information after questioning (Yield 1) and made more compliant responses after negative feedback (Shift). In Experiment 2, participants in the delayed condition internalized less material than those in the immediate condition. Conclusions : Different processes appear to underlie Yield 1 and Shift scores in the GSS2. The former may include both internalization of suggested materials and compliance, while the second appears to be mostly due to compliance with the interrogator. When administering the GSS2 in a forensic/legal context as vulnerability predictor for making false confessions, or proneness to develop false memories through the internalization of suggested material, including a source identification task may provide additional information on the type of coping style and memory characteristics of the examinee.     

Eyewitness Memory,Suggestibility, and Repeated Recall Sessions in Children with Mild and Moderate Intellectual Disabilities

This study of eyewitness memory questioned children with mild and moderate intellectual disabilities (ID) about a live staged event 1 day later and, again, 2 weeks later. Children with mild ID performed as well as typically-developing children of the same age in response to free recall instructions, and they were just as able as same age peers to resist misleading questions. However, they performed more poorly on general questions, probing for further information after free recall. The children with mild ID also changed their responses to specific questions more often in the repeated interview. The group of children with moderate ID showed markedly lower performance than peers of the same age on nearly every type of eyewitness memory question. Comparisons of the children with ID to mental age-matched peers indicated that performance was similar, although children with ID gave more information in response to free recall instructions and changed their answers in the repeated interview more often. Standardized measures of verbal memory (TOMAL) and suggestibility (Gudjonsson Suggestibility Scale) were modest to moderate predictors of eyewitness memory performance.     

The relationship between intelligence,memory and interrogative suggestibility in young offenders

Abstract

The study investigated the relationship between interrogative suggestibility and relevant cognitive factors as proposed by Gudjonsson and Clark (1986). The subjects were fifty-eight adolescent young offenders resident in a national children's centre with secure facilities. All were administered the Gudjonsson Suggestibility Scale and the Wechsler Intelligence Scale for Children. Interrogative suggestibility correlated negatively with intelligence and recall memory. Limited evidence of range effects was found for the relationship between intelligence and suggestibility, but not for the relationship between recall memory and suggestibility.     

Verbal ability,depression, and anxiety as correlates of interrogative suggestibility in children exposed to life adversities

In the present study, we assess the relationship of verbal ability, depression, and anxiety with interrogative suggestibility (IS) in a sample of children exposed to life adversities and requiring assistance by the governmental social service (n?=?39), as compared with normal controls (n?=?36). The Gudjonsson Suggestibility Scale 2 was administered to participants in a battery including measures of memory capacity, verbal ability, depression, and anxiety. The finding showed that the ‘assisted’ children were significantly more suggestible than the controls with a large effect size for Yield 1 and total suggestibility. In addition, IS was negatively correlated with verbal memory and ability, but not with anxiety and depression. The two groups differed as to their level of total suggestibility after adjusting for verbal ability. Implications of the findings for the forensic assessment of vulnerable subjects are discussed.     

The Portuguese adaptation of the Gudjonsson Suggestibility Scale (GSS1) in a sample of inmates

This paper comprises two studies which address the validity of the Portuguese adaptation of the Gudjonsson Suggestibility Scale, GSS1. In study 1, the means and standard deviations for the suggestibility results of a sample of Portuguese inmates (N = 40, M_age = 37.5 years, SD = 8.1) were compared to those of a sample of Icelandic inmates (Gudjonsson, 1997; Gudjonsson & Sigurdsson, 1996). Portuguese inmates' results were in line with the original results. In study 2, the means and standard deviations for the suggestibility results of the sample of Portuguese inmates were compared to those of a general Portuguese population sample (N = 57, M_age = 36.1 years, SD = 12.7). The forensic sample obtained significantly higher scores in suggestibility measures than the general population sample. ANOVA confirmed that the increased suggestibility in the inmates sample was due to the limited memory capacity of this latter group. Given that the results of both studies 1 and 2 are in keeping with the author's original results (Gudjonsson, 1997), this may be regarded as a confirmation of the validity of the Portuguese GSS1.     

Further data on interrogative suggestibility and compliance scores following instructed malingering

Purpose. This study examined whether people can successfully feign high levels of interrogative suggestibility and compliance as measured by the Gudjonsson Suggestibility Scales (GSS) and the Gudjonsson Compliance Scale (GCS) when given instructions to malinger. Methods. Participants (N = 90) were randomly allocated to one of three groups that received: (1) instructions to give into leading questions in order to look vulnerable to suggestions, (2) instructions to be compliant with the examiner, or (3) the standard GSS/GCS instructions. Results. One of the two malingering instructions led to modestly elevated scores of total suggestibility, while subscales remained unaffected. In contrast, both malingering groups obtained highly elevated compliance scores. Conclusions. These findings suggest that heightened suggestibility is rather difficult to malinger, thereby confirming the reliability of the GSS. On the other hand, it might be easier to malingering compliance as measured with the self‐report GCS.     

How well does the Gudjonsson Suggestibility Scale for Children,version 2 predict the recall of false details among children with and without intellectual disabilities?

Purpose . This study explored the effectiveness of the Gudjonsson Suggestibility Scale for Children, version 2 in predicting the tendency of older school‐aged children (with and without intellectual disabilities) to generate errors in an independent suggestibility paradigm. Method . Sixty‐nine children with an intellectual disability (aged 9–14 years) and 50 mainstream children matched for chronological age participated in a 30‐minute magic show that was staged at their school. Three days later, the children participated in a separate biasing interview that provided seven true and seven false details about the magic show. The following day, the children participated in a second interview where they were required to recall the magic show in their own words and answer a series of cued‐recall questions. Between 1 and 2 weeks later, the children were administered the Gudjonsson Suggestibility Scale‐2 (GSS‐2). Results . While there was no significant association between performance on the GSS‐2 and the independent suggestibility paradigm for the children with an intellectual disability, the chronological age‐matched children's yield scores predicted their reporting of both false‐new details and false‐interviewer suggestions for the independent event. Conclusion . When predicting children's recall of false details, the GSS‐2 appears to be more useful with mainstream school‐aged children compared with children who have an intellectual disability.     

Shortened versions of the Gudjonsson Suggestibility Scale meet the standards

Purpose. The Gudjonsson Suggestibility Scale (GSS; Gudjonsson, 1984, 1997 ) is a well‐established forensic tool for measuring interrogative suggestibility. However, one restriction of this tool is that it requires an extensive testing procedure. The present study examined whether shorter versions of the GSS yield similar results as the original GSS procedure. Methods. One group (N=20) was given a shortened version of the GSS that consisted of an immediate recall test and the specific questions. GSS scores in this group were compared with those in a group (N=20) that had the standard procedure which includes a retention interval and immediate and delayed recall tests. A third group (N=20) was administered a shortened procedure in which the 20 GSS questions immediately followed the GSS story. In the fourth group (N=20), participants were given the retention interval, but no recall tests were administered. Results. ANOVA showed no differences in GSS scores amongst the four groups. Post hoc power analyses indicated that these non‐significant findings were not the result of a power problem and that larger sample sizes are expected to yield comparable results. Further analyses showed that neither the retention delay nor the recall tests affected suggestibility scores. Conclusions. These results suggest that shortened procedures for administering the GSS may be employed in situations where time is a key factor.     

THE RELATIONSHIP OF ALCOHOL WITHDRAWAL SYMPTOMS TO SUGGESTIBILITY AND COMPLIANCE

People who are experiencing alcohol withdrawal are disadvantaged in terms of their ability to cope with leading questions and interrogative pressure (i.e. interrogative suggestibility). What had not been studied previously was the relationship of the severity of alcohol withdrawal symptoms with suggestibility and compliance. Suggestibility and compliance scores, obtained during the first week of hospital admission, were correlated with the severity of alcohol symptoms measured on a daily basis over a 7‐day period in a group of 393 patients attending treatment for alcohol abuse problems. Separate analyses were performed for the male and female patients. Significant gender differences emerged. Among males, alcohol withdrawal symptoms correlated positively with suggestibility and compliance across days. In contrast, among the females alcohol withdrawal symptoms were not significantly correlated with suggestibility and compliance, but were related to confabulations in memory recall. The findings suggest that in relation to psychological vulnerabilities during questioning, alcohol withdrawal symptoms are associated with different psychological factors in males and females. The findings have implications for the potential unreliability of information obtained from people interviewed during alcohol withdrawal.     

The relationship of suggestibility and compliance with self-deception and other-deception

The aim of the study was to assess the relationship of interrogative suggestibility and compliance with other-deception and self-deception. The Gudjonsson Suggestibility Scale (GSS 1), the Gudjonsson Compliance Scale (GCS), the Eysenck Personality Questionnaire (EPQ), and the Other and Self-Deception questionnaires (ODQ and SDQ) were administered to 237 Icelandic prison inmates. Other and Self-Deception scores did not correlate significantly with suggestibility and compliance, whereas significant correlations were found between the deception scores and EPQ Psychoticism, Neuroticism and Lie scores. The EPQ Lie score explained about one-third of the variance in the ODQ score, suggesting a considerable overlap between the two impression management measures. The most important practical implication of the findings is that self-favouring bias does not appear to significantly influence suggestibility and compliance scores.     

Testing eyewitness memory

This study identifies a memory-testing procedure that is relatively resistant to the documented effects of suggestibility on eyewitness memory. Most studies on suggestibility have used averbal recognition memory test in which the alternative test items are sentences, each to be verified as true or false regarding an originally viewedvisual sequence. In this study, participants were tested with either the verbal recognition memory test typical of studies demonstrating the eyewitness suggestibility effect or a visual recognition memory test. The typical eyewitness suggestibility effect resulted in the verbal test condition. However, with the visual recognition memory test, the hit rates did not significantly differ between the control and misled conditions. Thus, in testing memory for a visual event, a visual recognition memory test is more resistant to the influences of suggestibility than is a verbal test. These results suggest that the original item is preserved in memory, not overwritten by the misleading information. Accordingly, with a visual recognition memory test, the original information is more likely to be recovered with a visual recognition memory test than with a verbal one.     

The Role of Suggestibility in Determinations of Miranda Abilities: A Study of the Gudjonsson Suggestibility Scales

Traditionally, high levels of suggestibility have been widely assumed to be linked with diminished Miranda abilities, especially in relationship to the voluntariness of waivers. The current investigation examined suggestibility on the Gudjonsson Suggestibility Scales in a multisite study of pretrial defendants. One important finding was the inapplicability of British norms to American jurisdictions. Moreover, suggestibility appeared unrelated to Miranda comprehension, reasoning, and detainees’ perceptions of police coercion. In testing rival hypotheses, defendants with high compliance had significantly lower Miranda comprehension and ability to reason about exercising Miranda rights than their counterparts with low compliance. Implications of these findings to forensic practice are examined.