The current programme of research had two main aims. First, in response to a gap in the literature on implicit measures in the context of psychosis, the current thesis sought to determine the utility of the Implicit Relational Assessment Procedure (IRAP) in this regard (N=309). We opted to focus specifically on the IRAP because of the level of concept precision it has demonstrated previously in the context of depression, cocaine dependence and OCD. Second, we used the IRAP as a step toward bringing a broad, functional approach to understanding psychosis, by focusing on the specific phenomenon of voice hearing. On this path, we created a taxonomy of the critical features of voice hearing as identified in the literature (i.e. stigma, locus of control, evaluations of voices, normality of voices, acceptance of self and other people hearing voices, fear of voices, valence and acceptance of voices). It was our hope that outcomes in this regard might lead us one step closer to a functional analytic understanding of the psychological nature of the experience of hearing voices. The experimental designs and analytic strategies adopted in Experiments 1-3 were identical. All three studies involved control (non-voice hearing) participants completing one IRAP each. The six IRAPs employed across the studies had broadly similar trial-types that juxtaposed positive versus negative evaluations of various aspects of psychological suffering. Participants subsequently completed a battery of explicit measures which assessed psychological well-being and stigma. Experiment 1 investigated the broad stigmatisation of psychological suffering using three IRAPs (n=36 undergraduate students), namely a Depression IRAP, an Anxiety IRAP and a Mental Illness IRAP. Participants were assigned into one of the three IRAPs, in which they were required to relate suffering-as-positive and suffering-as-negative on alternative trial blocks. Overall, the three IRAPs in Experiment 1 produced similar patterns of responding, which, surprisingly, indicated that all three forms of suffering (i.e. depression, anxiety and mental illness) were implicitly evaluated as both positive and negative. Similarly, all three IRAPs also showed significant pro-normality effects, although an anti-normality effect was recorded on the Mental Illness IRAP. The correlations revealed that IRAP trial-types predicted some aspects of explicit. For example, positivity toward depression correlated with low anxiety and low stigma, while negativity correlated with stigma. Positivity toward normality also correlated with high distress. And pro-normality effects in the context of anxiety correlated with overall distress. Overall, these correlations suggested that psychological distress impacts upon implicit evaluations to suffering and normality, and thus, may influence stigmatisation. Experiment 2 investigated the locus of control in psychological suffering versus illness (n=40). We developed two IRAPs, namely an Illness IRAP and a Weakness IRAP, to investigate internal versus external locus of control. Participants were assigned into one of the two IRAPs, in which they were required to relate suffering-as-positive and suffering-as-negative on alternative trial blocks. Both IRAPs produced similar patterns of responding, in which suffering (as illness or weakness) was evaluated as both positive and negative. Significantly strong positivity toward health was also recorded. The correlations revealed that positivity toward health correlated with high stigma on the Illness IRAP and high distress and stigma on the Weakness IRAP, on which negativity toward weakness also correlated with low distress. Experiment 3 investigated stigmatisation of hearing voices by a non-voice hearing student population using a simple Evaluation IRAP (n=29). On the IRAP, participants were required to relate voices-as-positive and voices-as-negative on alternative trial blocks. Overall, hearing voices was positively evaluated. Surprisingly, implicit positivity toward voice hearing correlated with high stigma, but again negativity correlated with high distress. Experiment 4 investigated potential changes in implicit stigmatisation or fear of hearing voices in non-voice hearing student participants subjected to a hearing voices simulation and a Fear IRAP (N=28). On the IRAP, participants were required to relate voices-as-positive and voices-as-negative on alternative trial blocks. Subsequently, participants were exposed to a hearing voices simulation procedure, prior to completing the IRAP for a second time. Participants subsequently completed a battery of explicit measures. At baseline and post-simulation, hearing voices was implicitly evaluated as both positive and fearful, although positivity toward voices reduced and fear increased after the simulation. Experiment 5 investigated implicit evaluations of non-clinical voice hearers’ and non-voice hearing control participants (all from a student population) using the Normality IRAP (n=36). On the IRAP, participants were required to relate voices-as-normal and voices-as-abnormal on alternative trial blocks. Participants subsequently completed a battery of explicit measures. Both non-clinical voice hearers and controls implicitly evaluated voice hearing as normal, and most importantly this effect was stronger for the voice hearers. Evaluations of voice hearing as normal correlated with behavioural engagement with voices as well as voice acceptance, while evaluations of voice hearing as abnormal correlated with high voice benevolence and high emotional engagement with voices and other psychotic-like experiences. Experiment 6 investigated the potential stigma of non-voice hearing participants and non-clinical voice hearers (all from a student population) toward voices as heard by the self (i.e. “If I heard voices”) and others (i.e. “If Other People heard voices”), and presented these two groups with a Self IRAP and an Others IRAP (n=48). Participants were assigned into one of the two IRAPs, in which they were required to relate voices-as-positive and voices-as-negative on alternative trial blocks. Participants subsequently completed a battery of explicit measures. Overall, hearing voices was evaluated positively by both groups on both IRAPs, although control participants were more positive on the Self IRAP. Interestingly, control participants were also more fearful on the Self IRAP, while the voice hearers were more fearful on the Others IRAP. Furthermore, voice hearers who were less positive on the Self IRAP correlated with high depressive psychotic-like symptoms and those who were fearful on the Others IRAP correlated with low voice acceptance. Experiment 7 attempted to explore implicit evaluations by using a more broadly fear-based IRAP, and most notably by now including our first recruitment of clinical voice hearers (n=37). Clinical voice hearers were recruited from a psychiatric facility and the non-clinical voice hearers and non-voice hearing controls were recruited from a student population. On the IRAP, participants were required to relate voices-as-positive and voices-as-negative on alternative trial blocks. Participants subsequently completed a battery of explicit measures. As expected, controls and non-clinical voice hearers showed implicit positivity and only marginal fear on the IRAP. But, most interestingly, the clinical voice hearers were much less positive in this regard. The correlations also showed that implicit fear correlated with higher voice severity. Once again, these effects supported those we had observed previously, but critically showed that the clinical voice hearers were implicitly less positive than the other two groups. Experiment 8 sought to further explore the potentially different reactions of clinical and non-clinical groups to voices, and in doing so we attempted to parse out emotional versus behavioural responses (n=55). In a Valence IRAP, participants were required to relate voices-as-positive and voices-as-negative on alternative trial blocks, and an Acceptance IRAP required to relate avoidance and acceptance of positive and negative voices on alternative trial blocks. Clinical voice hearers were recruited from a psychiatric facility and the non-clinical voice hearers and non-voice hearing controls were recruited from a student population. Participants subsequently completed a battery of explicit measures. As expected, controls evaluated voice hearing positively on the Valence IRAP, but for the first time, non-clinical voice hearers showed implicit negativity, and we had now recorded for the second time, negativity by the clinical voice hearers. Clinical and non-clinical voice hearers implicitly accepted positive voices and avoided negative voices. Furthermore, acceptance of positive voices correlated with high psychological inflexibility, and acceptance of negative voices correlated with overall voice acceptance. This study provided the level of psychological precision that we had been working towards using the IRAP throughout the thesis. The refinements and the systematic manipulations of the IRAP have led us steadily from simple assessments of valence by control participants to distinguishing between the different types of reactions that different groups of voice hearers might have toward different types of voices, which largely coincides with the psychosis literature and thus suggests the validity of the IRAP within this domain.