Inflated perceptions of responsibility and obsessive–compulsive symptoms

Abstract

In Salkovskis' [Salkovskis, P. (1985). Obsessional–compulsive problems: a cognitive-behavioural analysis. Behaviour Research and Therapy, 28, 571–588; Salkovskis P. (1989) Cognitive-behavioural factors and the persistence of intrusive thoughts in obsessional problems. Behaviour Research and Therapy, 27, 677–682] cognitive model for obsessive–compulsive disorder (OCD), inflated perception of responsibility is highlighted as the critical feature that maintains the disorder. In the current study, the relationship between responsibility and obsessive–compulsive (OC) symptoms was examined. Specifically, three measures of responsibility were evaluated for their psychometric properties. These measures were then used to test Salkovskis's model by examining the relationships among pervasive responsibility, automatic thoughts related to causing harm, and OC symptom severity. Findings provide partial support for the validity and reliability of the measures and for the model as a whole. Based on regression analyses, data support that pervasive responsibility significantly contributes to the prediction of OC symptoms. Furthermore, this relationship appears to be mediated by automatic thoughts related to causing harm in OCD contexts.

Introduction

In his cognitive model of obsessive–compulsive disorder, Salkovskis, 1985, Salkovskis, 1989proposed that the root of obsessional pathology lies in the presence of schemas and negative automatic thoughts that relate to an exaggerated perception of responsibility. Specifically, stimuli are filtered through schemas dominated by fear of causing harm, which trigger automatic thoughts of potential harm in specific situations (cf. Rheaume et al., 1994). These thoughts, in turn, lead to distress and attempts to neutralize the potential for harm, that is, to compulsions, which are intended to make things right or to avert the possibility of being responsible for some feared outcome. Although this model makes sense on a theoretical level, empirical support for the full model is lacking.

The presence of inflated responsibility in people with OC symptoms has been examined with inconclusive results. Findings from several empirical investigations support the relationship between responsibility and OC symptom severity in nonclinical populations (e.g. Frost & Steketee, 1991; Freeston et al., 1991; Clark & Purdon, 1992; Freeston et al., 1992Steketee & Frost, 1994Rheaume et al., 1995). For example, Freeston et al. (1992)administered the cognitive intrusions questionnaire to college students and found responsibility, in conjunction with guilt and disapproval, to be a significant predictor of OC symptoms, accounting for 5% of the variance in compulsive activity scores. In another study (Rheaume et al., 1995), college students completed the authors' responsibility questionnaire and the Padua inventory (Sanavio, 1988). Regression analysis showed that responsibility accounted for up to 88% of the variance in OC symptoms. Furthermore, demonstrating support for the importance of responsibility in a sample of OCD patients, Lopatka & Rachman, (1995)found that both distress and urges to perform compulsions decreased when responsibility was reduced.

In contrast to these findings, other empirical investigations have found no association between OCD and responsibility (Rheaume et al., 1992; Frost et al., 1994; Rachman et al., 1995). Differences in findings may be attributable to differences in measurements used and subsequent distinctions in definition. For instance, in their assessment of responsibility, Frost et al. (1994)used a subscale of the Jackson personality inventory, which appears to specifically assess social responsibility (e.g. “I contribute to charity regularly”). In another study, Rachman et al. (1995)administered the responsibility appraisal questionnaire (RAQ), which defines responsibility not only in terms of causing harm to others, but also in terms of taking responsibility for social etiquette, and welcoming challenges for assuming responsibility over other people and things.

The conceptualization of responsibility that appears to show the most promise in demonstrating a clear association with OCD is one that defines responsibility in terms of belief in one's power to cause harm. Studies that have employed this definition have yielded the largest effect sizes for responsibility. For example, Rheaume et al. (1995)found that this conceptualization of responsibility predicted an additional 88% of the variance in scores on a measure of OC symptoms, controlling for the effects of perceived severity and probability of consequences. Rheaume et al. (1994)describe responsibility as the “belief that one has pivotal power to provoke or prevent subjectively crucial negative outcomes that may be actual or real world or moral” (p. 266). Salkovskis (personal communication, October 6, 1996) clarifies this definition by explaining that such influence is viewed as not only present, but it `the straw that broke the camel's back'. In other words, the person believes that harm would not emerge without his or her influence.

In sum, the literature contains some evidence of a link between responsibility and OC symptoms, particularly when responsibility is defined as belief in one's power to cause harm and is specifically anchored to OCD themes. Conceptually, then, this definition appears to fit into the model of Salkovskis, 1985, Salkovskis, 1989at the automatic thought level, the sense of responsibility triggered by specific OCD cues. Thus far, no studies have examined the relationship between OCD and responsibility in terms of causing harm in general, the more pervasive schema. In other words, no research has addressed whether exaggerated responsibility present in a variety of non-OCD contexts is also an important predictor of symptoms. Consequently, the absence of these data represents a gap in the empirical support for Salkovskis's model. The first purpose of the present investigation was to examine the relationship between pervasive responsibility and OC symptom severity.

Some researchers and clinicians have proposed that responsibility is more salient for certain types of OC symptoms, that is, checking as opposed to cleaning (Rachman & Hodgson, 1980; Lopatka & Rachman, 1995; Rheaume et al., 1995; van Oppen et al., 1995; Rachman, 1998). For example, Rachman (1998)anecdotally reported that checkers tend to have difficulty accessing anxiety during exposure when in the presence of a clinician, but not while alone in their homes, where responsibility is more salient. That this difference is less noticeable among cleaners, Rachman suggests, serves as an indication that responsibility may play a less significant role for this group. However, there is an alternative explanation for this finding. The presence of another may reassure a checker because the observer could confirm or disconfirm that an action was carried out. On the other hand, for cleaners, in the case where the action is infection by presumably invisible germs, an outside observer could offer no reassurance because he or she would not be able to see whether such an event occurred. Responsibility may simply be consistently of concern for cleaners, whereas, for checkers, there are periods of relief when others are present to diffuse that sense of responsibility. The question of whether responsibility is of greater importance for checking rituals rather than cleaning has never been examined empirically. Thus, a second purpose of this investigation was to test the hypothesis that responsibility is of particular importance to checking symptoms versus cleaning symptoms.

The final goal was to provide an initial test of the reliability and validity of a preliminary version of a new responsibility measure, a subscale of the obsessional beliefs questionnaire. This measure is under development by an international group of OCD experts (Frost et al., 1997) who hope that their new measure will serve as the standard in the field. Evaluation of the psychometric properties of this scale is needed.