The psychology of interrogative suggestibility: A vulnerability during interview

https://doi.org/10.1016/j.paid.2010.06.005Get rights and content

Abstract

This study uses structural equation modelling to investigate the psychological mechanism underpinning interrogative suggestibility on the Gudjonsson Suggestibility Scale (GSS). It considers the relationship between neuroticism (vulnerability especially) and compliance within the Five-Factor personality model, fearful avoidant attachment (FAA), the experience of intense negative life events (iNLE) and interrogative suggestibility. Each participant completed the GSS 1, the Life Events Questionnaire, the Relationship Scale Questionnaire, and the NEO Personality Inventory-Revised. Findings show that: (i) vulnerability and FAA correlate; FAA patterns indirectly affect misinformation acceptance both in the absence of pressure (Yield 1) and in response to pressure (Yield 2) through iNLE. (ii) FAA patterns and compliance indirectly affect sensitivity to interrogative pressure (Shift scores) through iNLE. An endogenous tendency towards distress, FAA, and compliant tendencies (with respect to Shift scores) may be the basis of individual differences in interrogative suggestibility. This could manifest as false statements and inconsistencies in answer during questioning.

Introduction

Interrogative suggestibility can be a serious psychological vulnerability during questioning (Gudjonsson, 2003). Across the academic and applied forensic setting the Gudjonsson Suggestibility Scale (GSS; Gudjonsson, 1997) is used to measure interrogative suggestibility: (i) Yield 1 scores measure misinformation acceptance in response to the pressure associated with questioning; (ii) Yield 2 scores measure misinformation acceptance in response to negative feedback; and (iii) Shift scores measure a tendency to change [initial] answers in response to negative feedback.

The Gudjonsson and Clark (1986) model has been the long-established theoretical framework explaining suggestibility. It specifies and focuses on factors that, during interview, appear central in encouraging suggestibility (e.g. heightened uncertainty, expectations of success, the use of negative feedback, and the presence of rapport). According to the model, suggestible responding is a result of an individual’s cognitive appraisal of the interview situation. The model though seems not to account for the background factors, of suggestible interviewees, which may culminate in: (a) heightened uncertainty and expectations of success which, in response to the pressure associated with questioning and external negative feedback, may in some cases lead to (b) individual differences in suggestibility on the GSS (see Gudjonsson, 2010). Research into this may enhance understanding of the basis of suggestible behaviour.

Research over the past few years has uncovered a consistent link between the reporting of intense negative life events (NLEs) and interrogative suggestibility on the GSS (Drake, 2010, Drake et al., 2008). Since this time further studies into why experiencing intense adversity may encourage suggestibility have been conducted. Drake et al. (2008) correlated self esteem scores with NLE and GSS scores finding no significant relationship. The significant relationship between NLE and GSS scores also persisted when memory recall ability was controlled for. Later work shows the effect of neuroticism, compliance and (again) NLE on Yield 2 and Shift scores, but fails to explain a significant proportion of the variance in Yield 1 scores (although neuroticism and compliance were included in the best fitting model; Drake, 2010). These studies suggest that neuroticism, compliance, and the experience of adversity seem to contribute to suggestible behaviour, but the lack of significant results with respect to understanding suggestibility on the Yield 1 subscale of the GSS implies the presence of other factors (not yet accounted for) that, together with the aforementioned variables, may help explain the variance in suggestibility scores across all three GSS subscales.

Neuroticism seems related to stress generation and a negative perception of situations (Safford, Alloy, Abramson, & Crossfield, 2007). This negative perception is commonly found in individuals scoring high on attachment anxiety and avoidance (Bowlby, 1988). These factors may then cause such individuals to have a lesser resilience to life stress (Beasley, Thompson, & Davidson, 2003) and to perceive events as more intensely negative (DeWitte and De Houwer, 2008, Gentzler and Kerns, 2006). Scarr and McCartney (1983) suggested that these traits may also draw out negativity from the social environment. Problematic dyadic interactions may therefore occur (Cyranowski et al., 2002). Vulnerable interviewees may, as a result, perceive the interviewer and the GSS task more negatively leading to a greater sensitivity to the pressure associated with questioning as well as to external negative feedback (Baxter et al., 2003, Gudjonsson and Clark, 1986). Fearful avoidant attachment patterns may also directly influence the acceptance of misleading information, due to such interviewees’ desire to avoid conflict and their eagerness to please (Muller, 2009).

Vulnerability is defined as a general susceptibility to stress and is a particular manifestation of neuroticism (Costa & McCrae, 1992). Vulnerability may be the aspect of neuroticism that relates to both attachment anxiety and avoidance, encouraging the reported experience of more intensely negative events. During interview this experience of intense negative life events may increase uncertainty, expectations of success, and Yield 1 scores. Fearful avoidant attachment patterns may also affect Yield 1 scores directly. In response to external negative feedback, the experience of more intense adversity might lead to greater sensitivity to that negative feedback, resulting in heightened uncertainty, expectations of success and higher Yield 2 and Shift scores.

Research suggests though that trait compliance, as well as neuroticism, fearful avoidant attachment patterns and the experience of intense adversity, may also have an effect on Shift scores (Drake, 2010). Trait compliance is defined as an established response to (expected) interpersonal negativity or conflict (Costa & McCrae, 1992). Neuroticism and attachment anxiety and avoidance encourage psychological distress and are related to compliance (Gudjonsson et al., 2008, Safford et al., 2007). Compliance may have developed as a coping method in the face of more intensely negative events. An established negative mindset may generate uncertainty, expectations of success, and cause answer shifting in response to negative feedback (Gudjonsson & Clark, 1986). The direct relationship between interrogative suggestibility and both compliance and neuroticism is also discussed in previous literature (see Gudjonsson, 2003). Vulnerability and compliance may thus also have a direct effect on Shift.

Based on the aforementioned theory three models are predicted:

Vulnerability (V) and fearful avoidant attachment (FAA; comprising high attachment anxiety and avoidance) are expected to be significantly positively correlated and will exert significant indirect effects Yield 1 through NLE (the mediator). Fearful avoidant attachment is predicted to also directly affect Yield 1.

It is expected that the model containing V, FAA, NLE and Yield 1 should provide the most satisfactory fit to the data (compared with the model including the neuroticism domain (N), FAA, NLE and Yield 1 scores).

V and FAA are expected to correlate in the positive direction. V and FAA are expected to exert significant effects on NLE; NLE will in turn significantly influence Yield 2.

It is expected that the model containing V, FAA, NLE and Yield 2 should provide the most satisfactory fit to the data (compared with the model including N, FAA, NLE and Yield 2 scores).

V, compliance (C) and FAA are predicted to be positively related and to exert indirect effects (through NLE) on Shift scores. V is predicted to also affect Shift scores directly.

It is expected that the BM containing V, C, FAA, NLE and Shift should provide the most satisfactory fit to the data (compared with the BMs including N, C, FAA, NLE and Shift scores).

Section snippets

Participants

The sample consisted of 120 participants, 94 females and 26 males (mean age = 19.35 years, standard deviation = 1.41, range = 18–26). Participants are an opportunity sample, recruited through the experimental participation scheme within the School of Psychology. All were undergraduates within the School.

Memory recall

The GSS memory recall task is presented in the form of a narrative, which is made up of 40 small instances, occurring in a specific order. Each instance is scored as ‘successfully recalled’ if the interviewee is able to freely recall that instance. The words used (by the interviewee) to recall the instances need not be exactly as written in the narrative. The maximum free-recall score that can be achieved is 40, which would indicate that the interviewee has correctly recounted everything that

Procedure

Participants were administered the GSS 1 individually by an interviewer, trained by a Chartered Forensic Psychologist in the administration of the GSS. In between the immediate free-recall and the interview phase, each participant completed the LEQ, the RSQ, and the N and A domain of the NEO PI-R (allowing trait vulnerability and compliance to be measured).

Descriptive statistics

Table 1 presents the means and standard deviations for the observed variables. The normality of the data was checked, revealing univariate normality for the GSS and NEO PI-R scores. Measures of iNLEs showed positive skewness (iNLE; Skew Z = 2.09, and kurtosis Z = 6.39). As a result, the use of parametric statistics was abandoned in favour of the non-parametric alternatives.

Spearman’s rho correlations

Table 2 shows the zero order correlations between the measured variables. Results indicate significant positive correlations

Discussion

The aim of this study is to investigate the psychological mechanism underpinning suggestible behaviour. Some key findings emerged:

Limitations and conclusion

There are a couple of limitations to this study such that the parameter estimates and fit statistics should thus be considered conservative estimates at this stage:

  • (i)

    A limitation may be the gender skew of the current sample. This may have affected the current relationships between the variables. Studies investigating gender differences in suggestibility on the GSS 1, however, indicate non-significant differences (Gudjonsson, 2003, p. 379). This suggests that gender may not affect performance on

References (28)

  • J. Cohen et al.

    Applied multiple regression/correlational analysis for the behavioural sciences

    (2003)
  • P.T. Costa et al.

    Revised NEO Personality Inventory (NEO-PIR) and NEO Five Factor Inventory (NEO-FFI) Professional manual

    (1992)
  • J.M. Cyranowski et al.

    Adult attachment profiles, interpersonal difficulties, and response to interpersonal psychotherapy in women with recurrent major depression

    Journal of Social and Clinical Psychology

    (2002)
  • K.E. Drake et al.

    Interrogative suggestibility, self-esteem, and the influence of negative life events

    Legal and Criminological Psychology

    (2008)
  • Cited by (20)

    • Measuring interrogative suggestibility with the Italian version of the Gudjonsson Suggestibility Scales (GSS): Factor structure and discriminant validity

      2015, Personality and Individual Differences
      Citation Excerpt :

      Recent research has uncovered that the experience of negative life events impacts considerably upon the interviewee’s interrogative suggestibility (Drake, Bull, & Boon, 2008). The experience of negative life events mediates the effect of psychological vulnerabilities, such as neuroticism, on GSS scores (Drake, 2010). In addition, it resulted that anxiety is not directly associated with interrogative suggestibility, but only has a significant moderating influence on the effect of a negative cognitive interpretation of the interview situation.

    • The role of trait anxiety in the association between the reporting of negative life events and interrogative suggestibility

      2014, Personality and Individual Differences
      Citation Excerpt :

      A limitation with the existing literature into interrogative suggestibility is that: (a) it provides insight into a range of predictors of interrogative suggestibility, without showing which, overall, are actually the most crucial (see Gudjonsson, 2003, 2013); and (b) although a few studies have examined interaction effects (see Baxter, Jackson, & Bain, 2003), there is a tendency to focus on the effects of exogenous factors (the effect of the interview situation/conditions/pressure and/or interviewer demeanour). This could lead to misleading conclusions that it is the negative feedback (interrogative pressure) or negative interviewer demeanour that is most central to interrogative suggestibility (see Baxter et al., 2003) or that life adversity is a marker of sensitivity to pressure, when this may not be the case (Drake, 2010; Drake & Bull, 2011). Research into intrinsic susceptibility factors is therefore necessary to unpick and clarify the mechanism driving and the crucial predictors of interrogative suggestibility.

    View all citing articles on Scopus
    View full text