A Model of the Relapse Process in Sexual Offenders.

Theorists and researchers have been attracted by the view that effective treatment of sexual offenders ought to be based on a clear understanding of the process of relapse (Laws, 1995; Pithers, 1990; Ward, Louden, Hudson, & Marshall, 1995). It has been suggested that there are clear patterns evident in the behavior of sexual offenders that translate into distinct clusters of cognitive, affective, and behavioral offense variables (Ward, Louden, et al., 1995). Models of the relapse process set out to provide a rich description of the cognitive, behavioral, motivational, and contextual factors associated with a sexual offense (Ward & Hudson, in press). Theory at this level typically includes an explicit temporal factor and focuses on proximal causes or the how of sexual offending. For example, the degree that planning is present, and if so what kind and at what point in the offense chain, the particular offending style exhibited by an individual, the degree of violence involved, and so on all should be represented in a rich descriptive model of the relapse process.

Although the empirical evidence for the existence of multiple offense pathways is still somewhat preliminary, it is reasonable to conclude that there are distinct pathways associated with sexual offending and that offenders constitute a heterogeneous group (Marshall, 1996). Our research suggests that some sexual offenses are associated with self-regulatory failure, and others with careful and systematic planning, accompanied by positive emotional states (Ward, Louden, et al., 1995). In two previous empirical studies, we have used grounded theory, a qualitative method, to develop a model of the offense process (Ward, Fon, Hudson, & McCormack, 1998; Ward, Louden, et al., 1995). These studies provided evidence for the existence of diverse offense pathways containing a number of distinct phases and suggest that offenders vary in their goals, their capacity to plan offenses, and in the kinds of emotions they experience throughout the offense process. Therefore, it seems prudent to base treatment on a thorough understanding of the specific deficits and behaviors exhibited by individual offenders, rather than assume that all men follow the same pathway when relapsing (Ward & Hudson, in press).

Historically, the assumption that relapse constitutes a process or chain of behavior occurring across time has led to the adoption of relapse prevention as a treatment model in the sexual offending area (Pithers, 1990; Pithers, Marques, Gibat, & Marlatt, 1983). Pithers's influential relapse prevention model focuses on factors proximal to offending, describing the process as an affective, cognitive, and/or behavioral chain that culminates in the recurrence of sexually aggressive behavior (Pithers et al., 1983). Despite its enormous value in guiding treatment and research, both Pithers's and Marlatt's (Marlatt & Gordon, 1985) original relapse prevention model suffer from a number of conceptual and empirical problems. Both have been constructed from very different, and arguably incompatible, theoretical elements leading to conceptual confusion and redundancy. However, from a clinical perspective, the most serious shortcoming of Pithers's framework is that it does not cover all the possibilities involved in reoffending. His model emphasizes skill deficits as the major mediators of relapse and fails to cover situations in which individuals consciously decide to use drugs or engage in sexually abusive behavior (for a systematic critique of both Marlatt's and Pithers's perspectives, see Ward & Hudson, 1996).

Drawing on our empirical and theoretical work, we suggest that a comprehensive model of the relapse process needs to contain a number of pathways, preferably taking into account different types of goals (e.g., approach versus avoidance goals), varying affective states (initial and ongoing), and different types of planning. Second, it should provide mechanisms to integrate cognitive, affective, and behavioral factors as they relate to the offense process. Third, it should include an explicit temporal emphasis and be able to account for the dynamic nature of the offense process. Fourth, it needs to be able to account for the various phases, or milestones, of the offense process, at least as they are currently understood. This includes the influence of background factors, distal vulnerability factors, decisions that lead to high-risk situations, the initial lapse, the sexual offense, and the impact of the offense on subsequent offending. In addition, the psychological mechanisms that drive or inhibit the relapse process should be identified and described.

In this article, a self-regulation model of the relapse process is described that we suggest addresses the problems that have been identified in Pithers's work. This model also has the potential to provide a more comprehensive guide for clinicians. First, we briefly review relevant research and theory on self-regulation to provide a conceptual basis for the model. Second, we present our model of the relapse process, which contains nine phases and four pathways. Finally, we follow up the abstract description of the model with case examples of each of the four pathways.

SELF-REGULATION AND SEXUAL OFFENDING

Self-Regulation Theory

Self-regulation consists of the internal and external processes that allow an individual to engage in goal-directed actions over time and in different contexts (Baumeister & Heatherton, 1996; Karoly, 1993). This includes the monitoring, evaluation, selection, and modification of behavior to accomplish one's goals in an optimal or satisfactory manner (Thompson, 1994). Therefore, it is clear that self-regulation is not solely concerned with inhibiting or suppressing behavior but can include the enhancement, maintenance, or elicitation of behavior as well. On some occasions, the enhancement of emotional states (e.g., when steeling oneself to tackle a difficult situation such as writing an exam) or precipitation of activity are legitimate goals. Also, the maintenance of behavior might be warranted when it has proved effective in achieving desired goals.

Goals are key constructs in theories of self-regulation and function to guide the planning, implementation, and evaluation of behavior. In essence, goals are desired states or situations that individuals strive to achieve or to avoid and, as such, are important components of personality (Austin & Vancouver, 1996; Emmons, 1996). Although goals vary according to their degree of abstractness, they can also serve different functions. Cochran and Tesser (1996) make a useful distinction between acquisitional (approach) and inhibitory (avoidance) goals. Acquisitional goals concern the gaining or increase of a skill or situation, and essentially involve approach behavior. Failure to achieve such goals tends to be a graded occurrence and may function to increase a person's effort to succeed. Attention is focused on information indicating success, and therefore positive memories and cognitions are more likely to be experienced. By way of contrast, inhibitory goals are concerned with the decrease or inhibition of a behavior or situation, and involve avoidance behavior. Failure is usually construed in an all or nothing manner, and attention is focused on information signaling failure rather than success. Therefore, failure related memories or cognitions are more commonly experienced by individuals whose behavior is guided by inhibitory goals.

Inhibitory or negative goals are always more difficult to achieve, as there are many ways in which a person can fail to prevent an event or state from occurring (Wegner, 1994). Such a task demands considerable cognitive resources because it is necessary to monitor the environment for all types of potential threats. Therefore, when an individual is stressed or experiencing strong emotional states, self-regulation can be impaired relatively easily. In contrast, there may only be a single route to a desired goal, and the major drain on cognitive resources concerns the planning and implementation of the actions necessary to obtain the desired outcome (Emmons, 1996). From a self-regulatory perspective, such a task is more straightforward. The evidence suggests that individuals with avoidance goals experience higher levels of psychological distress than do those with approach goals (Emmons, 1996).

In their control theory of self-regulation, Carver and Scheier (1981, 1990) argue that goals are cognitive structures stored in memory in the form of behavioral scripts or knowledge. These cognitive representations contain information that enable individuals to interpret the actions of others and also to guide their own actions. Goals and the accompanying behavioral scripts are hierarchically organized and vary according to their degree of abstraction. For example, shoveling snow off the sidewalk; helping one's neighbors; or, at a very abstract level, trying to be a thoughtful and caring individual are all descriptions of related goals. The more abstract or higher level goals are translated into lower level behaviors and associated subgoals and ultimately physical actions.

Goals are also related to affective states, with achievement being associated with positive emotions and failure with negative emotions (Carver & Scheier, 1990). Negative affective states typically result in more extensive processing than positive states, probably because they indicate failure to achieve a valued goal (Carver & Scheier, 1990). Some goals can be directly activated by environmental factors and can result in automatic goal-directed behavior without the need for conscious decision making (Bargh & Barndollar, 1996). Once such goals are activated, they exert a direct influence on the processing of information and subsequent generation of behavior.

The self-regulation of behavior typically …