It is important to note the methodological limitations of the studies involved in this thesis. As previously discussed, an important limitation in this research program is the reliance of the first three studies on a student sample. The use of a non-clinical sample that shows cognitions and symptoms similar to those suffering from OCD made it possible to examine more complex relationships between constructs (such as moderation and structural relationships). However, the conclusions from these studies are limited as there may be questions regarding the extent of generalisability to clinical participants. The last study aimed to replicate previous findings with clinical samples. However, the number of participants in this study limited the ability to utilize sophisticated statistical methodologies and thus examine the complex relationships between the constructs. Future research would benefit from the use of a larger sample of clinical participants.
Another important limitation of this study is the reliance on correlational data. An experimental design examining the hypothesized relationships could lead to stronger causal conclusions. For instance, potentially useful experimental designs could identify participants' sensitive self domains (via interview or questionnaire measures) and ask participants to listen to idiosyncratic distressing intrusive thoughts following prompting of their "sensitive" or "non sensitive" domains (e.g., by visualizing being complemented or criticized in the "sensitive" or "non sensitive" domain). Measures of distress and other relevant variables (e.g., appraisals, urge to neutralize, etc.) could then be recorded. Such a design could allow stronger conclusions about the relationship between sensitivity of self and intrusive thoughts and OC phenomena.
It has been implied throughout this thesis that early attachment experiences result in particular internal representation of self and the world. These later increase the likelihood for developing OC related cognitive affective structures. Although there is considerable evidence to suggest that the relative stability of attachment representations (see Farley 2002 for meta analysis; also see Chapter 2 of this thesis), environmental changes may reduce such stability of attachment representations (e.g., Bowlby, 1973; Guidano & Liotti. 1983; Waters et al., 2000; also see Sroufe et al., 1999. Thus, as mentioned earlier (see chapter 6 of this thesis), the link found between adult attachment and OC phenomena may partly reflect the influence of negative life events. Alternatively, OC phenomena may increase the likelihood of developing negative adult attachment representations. Again, future research using longitudinal and experimental approaches may provide stronger evidence of causality between these constructs.
The present study did not examine perceptions of parenting. As mentioned previously (see chapter 2 of this thesis), sensitivity to the development of OC symptoms may be associated with specific parenting variables, which are not evaluated by existing parenting measures. The development of such a measure was beyond the scope of this thesis. However, an examination of the literature may give clues with regard to the parenting or other variables involved in the development of an individual's over-reliance on certain aspects of self (i.e., a self-concept comprising relatively few domains that are 'sensitive') coinciding with the belief that the world is controllable but threatening; such patterns of parenting could be specific to OCD, consistent with the specificity of self-structures to OCD. For instance, Rogers (1951) described the influence of parental conditions of approval and disapproval on their children's self-perceptions. Parental reactions and judgments establish 'conditions of worth' by which the child later attributes value (positive or negative) to certain aspects of self. Attachment theory also suggests that parenting plays an important (although not exclusive) role in the formation of self-worth and self-esteem through the construction of internal working models (Bowlby, 1988). The goodness of fit between the parent and the child and associated accessibility and responsiveness to the child's needs when stressed and anxious (e.g., when the child is in a novel and therefore potentially threatening environment) and the quality of attention that the child receives from their parents, play an important role in the formation of the sense of self and general perceptions of the world (Guidano & Liotti, 1983; Salkovskis et al., 1999).
It can be argued that parental availability, contingent on a child's performance and/or competence (related to moral and/or social domains), may lead to these domains being increasingly valued. Alternatively, parents may pay negative attention to the child and may even reject or punish the child when expected standards are not met in valued domains or when the child seeks access to parental attention in other domains (i.e., domains not valued by the parent). Thus, both positive and negative attention could reinforce the importance of the valued domains. In this way, self-perceptions that are over-reliant on specific domains of competence could be formed. In addition, parental rigidity and an inflexible attitude with regard to the child's performance in these domains may convey the belief that certainty and perfection are achievable and desirable (Guidano & Liotti, 1983; Salkovskis et at, 1999). Parental expression of excessive anxiety and worry will convey the message that the world is dangerous (Kyrios, 1998; Salkovskis & Forrester, 2002), enhancing perceptions of the world as a malevolent place (Janoff-Baulman, 1991). Thus, vulnerability to intrusive thoughts of a specific nature, together with a world-view consisting of both perceptions of threat and danger and a strong belief in the achievability and/or need for control is developed.
Such parenting styles are expressed in behaviors such as praising, helping, supporting, and spending quality time with the child, primarily when certain valued standards are met in specific domains. A relative lack of attention in areas not valued by the parent (but valued by the child) will result in the child developing generally low perceptions of self-worth. Parental attention (positive and negative), limited mainly to the specific domains valued by parents, leads to strong associations between feelings of self-worth and competence in these domains. The child perceives the parents as affectionate and loving (and, as result, themselves as worthy of love) only in the context of achievement in a limited number of domains. Because such standards are rigidly held/reinforced and are difficult (if not impossible) to attain, this compounds the child's general feelings of failure and low self-worth. Thus, investigating more specific parenting styles and their association with later dysfunctional views of self and the world may help develop parenting intervention that help in reduce vulnerability to OCD.
Finally, although the questionnaires in this study showed adequate reliability, some of the questionnaires showed lower reliability (e.g., the WAS) than other questionnaires (ECR, BDI, OBQ). Among other difficulties (e.g., lower validity), lower reliability of a measure may hinders detection of the expected effects. Future research may consider the development of more reliable measures for examining such constructs.