Measurement of Individual Differences in Children’s Suggestibility Across Situations Matthew H. Scullin, Tomoe Kanaya, and Stephen J. Ceci Cornell University The authors attempted to use scores on the Video Suggestibility Scale for Children (VSSC, M. H. Scullin & S. J. Ceci, 2001) to predict 50 preschool children’s performance during a field study in which they were interviewed suggestively 4 times about both a true event and a suggested event. Among the 25 children over age 4 years 6 months, tendencies on the VSSC to respond affirmatively to suggestive questions (“yield”), change answers in response to negative feedback (“shift”), and the sum of these (“total suggestibility”) were all related to lack of accuracy about the true event in the field study and to both accuracy and lack of accuracy about the suggested event. Results support a 2-factor model of suggestibility. Since Alfred Binet’s pioneering work on suggestibility nearly 100 years ago (Cunningham, 1988), psychological researchers have conducted numerous studies to identify external or “situa- tional” variables that affect children’s testimony (see Ceci & Bruck, 1993, 1995, for reviews). More recently, researchers have begun to examine the characteristics within children that cause some to be more accurate than others (for reviews see Bruck, Ceci, & Melnyk, 1997; Quas, Qin, Schaaf, & Goodman, 1997). Accord- ing to these reviews, there is a growing consensus that children’s suggestibility entails an interplay between individual characteris- tics and situational factors and that further research is needed to examine this relationship in greater detail. There are at least four reasons for examining the interplay between developmental, situational, and individual characteristics in detail: First, young children have been found to be more suggestible than adults or older children, with preschool children being the most suggestible (Ceci & Bruck, 1993). However, even among very young children there is still a great deal of variability, with some younger children being less suggestible than some older children. It is unclear how much of this variability is related to relatively stable individual differences among children of the same age—for example, in general cognitive functioning—and how much may be due to maturational differences in specific cognitive attainments such as theory of mind (Templeton & Wilcox, 2000; Welch-Ross, 1999), source monitoring ability (Lindsay, Johnson, & Kwon, 1991), or executive control (Perner & Lang, 1999). Second, there is a great deal of interest among lawyers, case- workers, and psychologists for a means of determining the degree to which a particular child is vulnerable to leading questions and social pressure (Quas et al., 1997). Because of the large number of young children entangled in the juvenile, family, and criminal justice systems as a consequence of abuse and neglect proceedings, acrimonious custody disputes, and Persons in Need of Supervision actions (Ceci & Friedman, 2000), it is useful to examine the role of individual differences in suggestibility. This is because inves- tigators are less interested in being told that young children are, on average, more suggestible than older children, and instead prefer to be told whether a particular child is suggestible. A valid measure of individual differences in suggestibility could help identify chil- dren who are prone to making erroneous reports following sug- gestions and provide important information about when special interviewing precautions should be taken. Third, forensic studies have found that adults who make false confessions score higher on a psychometric measure of interrog- ative suggestibility than do control groups (Gudjonsson, 1992b). This finding suggests that it may be fruitful to use a similar approach to study differences in suggestibility of children. Finally, the Video Suggestibility Scale for Children (VSSC), comparable in some psychometric respects to adult suggestibility scales (Gudjonsson, 1984, 1987), was recently developed (Scullin & Ceci, 2001). However, it is unclear whether the scale has construct validity. The study reported here is an attempt to estab- lish the VSSC’s construct validity by comparing participants’ performance with a variety of measures derived from a separate study in which the same children were interviewed suggestively about both real and imaginary events. In what follows, we briefly discuss the nature of suggestibility and how interviewer bias and suggestive interviewing techniques Matthew H. Scullin, Tomoe Kanaya, and Stephen J. Ceci, Department of Human Development, Cornell University. This research was supported in part by grants from the College of Human Ecology at Cornell University. We thank Angela Crossman, Livia Gilstrap, and Heather Warren for their contributions to this study, and Jersy Chen, Amar Gavhane, Stephanie Linkous, Jessica Passero, Nina Piccoli, and Lisa Tallon for their data collection and coding. We thank the teachers and parents at Ithaca Community Childcare Center and IACC in Ithaca, New York, Happy Way Childcare Center in Dryden, New York, and Newfield Day Care, Newfield, New York, for their participation in this research. Correspondence concerning this article should be addressed to Matthew H. Scullin, who is now at the Department of Psychology, West Virginia University, Morgantown, West Virginia 26506. E-mail: mhscullin@ mail.wvu.edu Journal of Experimental Psychology: Applied Copyright 2002 by the American Psychological Association, Inc. 2002, Vol. 8, No. 4, 233–246 1076-898X/02/$5.00 DOI: 10.1037/1076-898X.8.4.233 233