Centuries after it was first described, dissociation is one of the most poorly understood constructs in clinical psychology. Used interchangeably to signify either a symptom or an aetiological process, the term 'dissociation' has attracted multiple definitions incorporating a diverse range of experiences. The confusion conceptualising dissociation, as well as longstanding debates arising from significant theoretical and socio-historical challenges in the field, have contributed to the under-recognition of dissociative phenomena across clinical settings, and to difficulties researching the topic. As a result, the psychological mechanisms underpinning dissociation remain unknown. More precisely, to date, there are no cognitive models explaining the development or maintenance of dissociation that have been developed through extensive causal testing of proposed factors, and accordingly, there are no translational psychological treatments specifically targeting this construct. Furthermore, due to its importance in the field of trauma, dissociation has rarely been explored as a transdiagnostic experience in its own right, nor has there been full consideration of its influence across the spectrum of psychological distress. This omission may be particularly important in the field of psychosis, where there is increasing evidence that dissociation has robust links with positive psychotic symptoms, such as paranoia, delusions, and hallucinations. Dissociative experiences have even been implicated as a possible cause of such difficulties, as they create conditions ripe for uncertainty, misinterpretation, and distress. In the same way, dissociation has been suggested to sustain psychotic symptoms. Therefore, under-recognised in its own right, dissociation has until recently also been under-recognised in terms of its importance in psychosis, and is only now being explored as a potential novel treatment target. In order to overcome the legacy of decades of entrenched debates and confusion, new approaches to exploring dissociation are needed. Network analyses are based on network theory, which proposes that mental health problems can be understood in terms of patterns of activation between symptom 'nodes' connected in a network of conditional dependencies. Such analyses are gaining increasing popularity in mental health research and may be well-suited to providing fresh insights into dissociation as they allow the estimation and visualisation of inter-correlations between multiple variables at once. Some forms of network analysis may also benefit the field through the generation of novel testable hypotheses: directed acyclic graphs estimate the probable direction of effect between correlated variables using statistical probability and may therefore be particularly valuable in early-stage research. Aims The aims of this thesis are to clearly define the phenomenon of dissociation, demonstrate its importance in psychosis, and begin a scientifically robust interrogation of the construct which will inform the future development of a cognitive model. In order to achieve this, the first half of the thesis develops tools required for valid research from a cognitive perspective: a precise definition of the construct, and two self-report measures with which to assess dissociation and cognitive appraisals of dissociative experiences. In the second half of the thesis, state-of-the-art network analyses are used to locate dissociation within the wider context of mental health (including in relation to both common affective and psychotic experiences), and to begin determining the psychological factors underlying dissociation. Finally, key findings of the thesis are tested in the context of psychosis. Methods Data were collected from four participant cohorts: a small psychosis patient group (n=12, Chapter Two), two large online general population samples ('Mapping Dissociation in Mental Health': n > 8000, Chapters Three and Five; and 'Understanding Dissociative Experiences': n > 8000, Chapters Four and Six), and a large psychosis patient group ('Exploring Unusual Feelings', n=1038, Chapters Three, Four and Seven). Chapter Two uses a qualitative study of twelve NHS patients with psychosis diagnoses to better understand the lived experience, and hence describe the construct of, dissociation in psychosis. In Chapter Three, a systematic review of dissociative measures is carried out and triangulated with the results of Chapter Two in order to propose a novel theoretical framework which describes dissociation as it will be conceptualised for the purposes of this thesis. The theoretical framework is empirically tested via factor analysis of data from the Mapping Dissociation survey, generating a novel self-report measure of the new conceptualisation of dissociation. This measure is then validated using data from the clinical study Exploring Unusual Feelings. Chapter Four introduces a second psychometric tool: a measure of cognitive appraisals of dissociation developed with data from the Understanding Dissociative Experiences survey and validated using data from Exploring Unusual Feelings. Chapter Five explores the relationship of dissociation to other mental health presentations, including psychotic and affective symptoms. Data from Mapping Dissociation are modelled in two network analyses - a Gaussian graphical model and a Bayesian inference with directed acyclic graphs (DAGs) model. Chapter Six uses the same analytical methods to explore the relationships between dissociation and specific psychological factors in data from Understanding Dissociative Experiences. Psychological factors explored in this study include affect intolerance, perseverative thinking, anxiety sensitivity, alexithymia, and body vigilance. Finally, Chapter Seven tests the key findings from the previous two chapters in a large clinical patient sample (Exploring Unusual Feelings). Results Within the context of non-affective psychosis, dissociation is described by patients as having a negative impact on their relationships with other people, and on their existing mental health difficulties, mood, cognition, daily functioning, and sense of self (Chapter Two). As well as highlighting the significant difficulty inherent in describing their highly subjective nature, the results also reveal that a common denominator of participants' dissociative experiences is a felt sense of anomaly (FSA) where this is conceptualised as a subjective feeling of oddness or strangeness. Triangulation of these results with a systematic review carried out in Chapter Three leads to the proposition that FSA may take a number of 'forms' (e.g. unreality, unfamiliarity, detachment) and affect any of a range of 'domains' (e.g. perception of the external environment, experience of one's emotions). Empirically testing the proposed theoretical framework using factor analysis reveals seven key factors of FSA-dissociation: anomalous experiences of the self, body, and emotion, and altered senses of familiarity, connection, agency, and reality. The self-report scale developed in this chapter was found to be psychometrically robust in general population and clinical groups, and is therefore used as the primary measure of dissociation throughout the thesis. Chapter Four develops and validates a new measure of cognitive appraisals of dissociation and demonstrates its additional explanatory power above that of FSA-dissociation alone when predicting hallucinations and paranoia scores in a clinical group using regression analyses. These results, as well as the results of Chapter Five, support the hypothesis that dissociation has a causal influence on positive psychotic symptoms. In Chapter Five, network analyses find that dissociation is a probable cause of paranoia, grandiosity, and cognitive disorganisation. Dissociation is also found to have a robust relationship with hallucinations, but as with its relationship to depression and anxiety, the direction of influence is unclear. Network analyses in Chapter Six find direct relationships between dissociation and cognitive appraisals, safety behaviours, perseverative thinking, low general self-efficacy, and low alexithymia. Testing these candidates for potential maintenance mechanisms of FSA-dissociation in a clinical group (Chapter Seven) reveals direct relationships between dissociation and cognitive appraisals, safety behaviours, perseverative thinking, affect intolerance, and low alexithymia. The results of these network analyses also support the hypothesis that dissociation is related to lower wellbeing, and may cause psychotic symptoms: both paranoia and hallucinations had direct relationships with dissociation. Conclusions This thesis defines the construct of dissociation (as it is experienced in the context of non-affective psychosis) as having a core phenomenological experience of a felt sense of anomaly (FSA). Dissociative experiences of this kind are experienced as distressing, disruptive, and appear to be a probable cause of cognitive disorganisation, paranoia, and hallucinations. They may be experienced in relation to a range of 'domains' (e.g. one's body or emotions) and involve different 'types' of anomaly (such as unfamiliarity or detachment). Appraisals of such experiences typically include fears about social, biological, or psychological ill effects, as well as concerns about lack of control or agency. Such appraisals are important in the development and maintenance of dissociative experiences and may partially explain the effect of dissociation on psychotic symptoms.