The first step in studying a phenomenon such as stalking is to con¬duct descriptive and correlational studies that provide information on how and why stalking occurs. Another important area of study is the analysis of antecedents, or early signs of stalking, that might be used to pre¬dict the behavior in specific situations or in certain individuals. In this con¬text, early signs of stalking constitute either of the following: (a) problem behaviors in the social relationships of children or adolescents that could be viewed as early manifestations of obsessional following in a specific instance, or (b) childhood risk factors that suggest someone might stalk another person later in life. One reason for exploring the potential early signs of stalking in young people is that it broadens the application of mate¬rial presented thus far by raising awareness of those risk factors that can be targeted for treatment. In this way, preventive programs can be implement¬ed early on so that the risk of stalking might be reduced. A final reason for examining potential risk factors for stalking is that some problem behaviors in childhood that overlap with stalking, such as bullying and sexual harass¬ment, have been more extensively studied and can provide direction for the design of preventive programs for stalking and obsessional following.
STALKING IN CHILDREN AND ADOLESCENTS
Before discussing these early risk factors, I offer a note of caution on the practical use of the material presented in this chapter. When informa¬tion is offered for one purpose, such as providing clinical and empirical hypotheses or outlining specific issues that can be addressed in treatment, there is a risk that such factors may be applied for less legitimate purpos¬es, such as identifying specific individuals for restrictive detention or con¬firming some preconceived notion about a person's guilt or innocence in a criminal case. The present chapter is offered as an overview of clinical observations, anecdotal evidence, and research findings on topics related to stalking that outline hypotheses for understanding how stalking may develop and what preventive strategies might be effective. The potential for misuse of the material presented herein arises if broad generalizations are drawn about a particular child or adolescent without confirming or disconfirming these hypotheses through the col¬lection of other data in a comprehensive psychological assessment. It is important to note that the process of clinical judgment and prediction may produce errors if clinicians fail to recognize sources of bias in judg¬ment or misapply heuristics, or simple rules for making clinical decisions. Garb (1998) provided a detailed discussion of biases that can occur in clinical judgment. Confirmatory bias occurs when clinicians look only for data that confirms their hypotheses but overlook the evidence that refutes their hypotheses. In the present context, the clinician must avoid focusing only on those specific behavioral signs observed in a particular youth that suggest a proclivity to become an obsessional follower without also con¬sidering evidence to the contrary. Hindsight bias is an error in judgment that suggests once a given event has occurred, the clinician believes that the event could have been predicted. This bias is often framed in the adage "hindsight is 20/20." Hindsight bias is more relevant to the issue of behav¬ioral prediction of violence and is addressed in the next chapter. Misestimation of covariance is an error in clinical judgment that occurs when clinicians fail to properly describe the relationship between two variables; that is, the clinician may recall times when one variable was present and another variable was also observed but does not recall times
THE EARLY SIGNS
when the two variables were unrelated. In the present context, misestima-tion of covariance is an important issue to recognize. For example, if a particular risk...