The impact of mood on the endorsement of rumination: an exploration of the association between cognitive biases and behaviour

Abstract

Reliable and robust methods of assessment are crucial to best practise within research and clinical settings. Although there are a number of benefits associated with the use of self-report questionnaires there is cause to examine how accurately they measure the frequency and content of mental processes. In particular, little attention has been paid to the potential influence of mood to effect endorsement behaviour. The existence of mood related variation in response styles would represent a major threat to the correct interpretation of self-report data. In clinical settings this threat is further increased due to samples of respondents who are expected to experience variations in mood across the course of therapy and relevant measurement points (e.g. pre-post treatment outcome measures). This dissertation sought to assess how mood may influence the endorsement of ruminative behaviours. The choice to examine rumination was threefold, rumination has developed as a key construct in understanding cognitive vulnerabilities to depression, it is solely assessed via self-report methodologies, and concerns have been raised that existing questionnaires are overly negative in tone. This is of particular significance given that there is a large body of literature demonstrating that depression is characterised by cognitive biases for negatively valenced stimuli. Such biases have been demonstrated in attention, memory, and interpretation -cognitive processes that are intrinsic to endorsement behaviour. If negative affect influences endorsement of ruminative behaviours, the aforementioned cognitive biases may play a role. This was explored over five studies. Results demonstrated that respondents are sensitive to question valence, and that subtle changes in wording can inflate item negativity. Furthermore, results indicated that experimentally induced negative mood increased perception of item negativity and endorsement of ruminative items. The final two studies highlight possible options for quantifying and minimising the effects of mood related changes in endorsement via the development and validation of a novel measure for endorsement bias, and support for the incorporation of informant-reports in the assessment process. Show more