Gudjonsson suggestibility scale (GSS): Further data on its reliability, validity, and metacognition correlates, The

Social Behavior and Personality, 1998 by Merckelbach, Harald, Muris, Peter, Wessel, Ineke, van Koppen, Peter J

This article presents two studies in which the psychometricproperties and validity of the Gudjonsson Suggestibility Scale (GSS) were further investigated. Results of the first study (N = 40) indicate that the GSS has reasonable internal consistency. Additionally, a modest, but significant test-retest stability was found for the GSS. As to the association between suggestibility and self-reported cognitive efficiency (i.e., metacognition measures), scores on the Yield dimension of the GSS were positively and significantly related to scores on the Dissociation Experiences Scale (DES), but not to scores on the Cognitive Failures Questionnaire (CFQ). In a second, experimental study (N = 53), evidence was found for the predictive validity of the GSS. In that study, subjects saw a slide series and were then confronted with leading questions about the critical (emotional) slide. In addition, they completed the Yield scale of the GSS. A small but significant correlation was found between subjects' Yield scores on the GSS and their susceptibility to leading questions about the slide series.

The Gudjonsson Suggestibility Scale (GSS) is an instrument that intends to measure individual differences in interrogative suggestibility (Gudjonsson, 1984a; 1992; 1997). More specifically, the GSS taps two different aspects of interrogative suggestibility, namely the tendency to give in to leading questions (Yield) and the tendency to shift responses under conditions of interpersonal pressure (Shift). In short, the GSS consists of a narrative paragraph that is read out to the subject, who then reports all he or she recalls about the story. Following this, the subject is asked a number of questions about the story, some of which are (mis)leading. Next, the subject is told in an authoritative manner that he or she has made a number of errors and must answer the questions for a second time. Yield refers to susceptibility to suggestive questioning, while Shift pertains to pressured suggestibility, i.e., the tendency to change answers as a result of social pressure.

Several studies have evaluated the psychometric properties of the GSS. In general, these studies found that the GSS Yield and Shift scales have sufficient internal reliability and test-retest stability (Gudjonsson,1992), although it is also clear that Shift scores critically depend on whether the pressured instructions are presented in a convincing manner. Furthermore, factor analytic studies have revealed a factor structure that fits the Yield and Shift dimensions (e.g., Gudjonsson, 1984a; 1992). In addition, there are field and case studies that have found evidence to support the validity of the GSS (see, for a review, Gudjonsson, 1992; see also, Gudjonsson, 1995). For example, in a study by Gudjonsson and Singh (1984), GSS scores of delinquent boys were found to be correlated with judgements of their suggestibility made by their teachers. As another example, Gudjonsson (1984b) reported that suspects who retract their confession ("false confessors") have higher GSS scores than suspects who consistently deny any involvement in the crime they are charged with ("deniers").

Secondly, a number of studies have examined psychological and psychopathological correlates of GSS Yield and Shift scores. By and large, this research has revealed a theoretically meaningful pattern of correlations. For example, age and memory are both negatively related to GSS Yield and Shift scores (e.g., Danielsdottir, Sigurgeirsdottir, Einarsdottir & Haraldsson, 1993; Gudjonsson, 1984a), while correlations with anxiety measures are consistently higher for Shift than for Yield suggestibility (Gudjonsson, 1988; but see Gudjonsson, Rutter & Clare, 1995).

The GSS is increasingly used for practical forensic purposes (e.g., Kassin, 1997). Forensic application of the GSS can be expected to increase even more now that trauma therapists (e.g., Brown, 1995) have recommended the GSS as an instrument for identifying patients who are vulnerable to develop pseudomemories.

The current article presents data from two studies in which the GSS was further investigated. Note that most work on the psychometric properties of the GSS has been done by Gudjonsson and co-workers (see, for a summary, Gudjonsson, 1992). The first study sought to replicate their basic findings. Thus, the first study was concerned with the psychometric properties (i.e., internal consistentcy and test-retest stability) of the GSS. In addition, this study examined the connection between suggestibility as indexed by the GSS and some widely used metacognition self-report scales, namely the Dissociative Experiences Scale (DES) and the Cognitive Failures Questionnaire (CFQ). These scales were included because they provide important information about how a person evaluates his or her own cognitive capabilities. One could argue that uncertainty about one's own cognitive efficiency (e.g., "memory distrust"; see e.g., Gudjonsson, 1992) contributes to suggestibility. The second study sought to validate the Yield scale of the GSS against an experimental standard. That is, GSS Yield scores of subjects were related to their tendency to give in to leading questions about a slide series.