Objective: This study aimed to test predictions made by the self-memory system (SMS) model (Conway & Pleydell-Pearce, 2000), extensions of Williams et al.’s (2007) “capture and rumination” (CaR) mechanism (Debeer, Hermans, & Raes, 2009) and control theories of rumination (Martin & Tesser, 1996, 2006) in a non-clinical sample to further understand the processes underlying overgeneral memory (OGM). It was hypothesised that (i) ruminating on unresolved goals, compared to thinking about resolved goals, would increase OGM, in participants reporting high levels of trait brooding and (ii) that this effect would be greater following goal-cues that are derived from goal cues rated as (a) more important compared to those rated as less important; (b) more progress-discrepant compared to those rated as less progress-discrepant; (c) more relevant to unresolved goals compared to those that are rated as less relevant to unresolved goals. Method: A between-subjects factor of condition (resolved versus unresolved goal-focus induction) and a within-subjects factor of time (pre- and post-manipulation Minimal Instruction Autobiographical Memory Test [MI-AMT; Debeer et al., 2009]) design was utilised with 75 undergraduate and three masters psychology students (86.3 % female [n = 65]; age, M = 20.2 years, range = 18-43, SD = 4.9) from the University of Exeter. The MI-AMT was used to measure autobiographical memory (AM) specificity before and following a manipulation whereby participants were randomly assigned to either a control condition in which participants focussed on a resolved goal or an experimental condition which was designed to induce rumination about an unresolved (i.e., self-discrepant) goal. MI-AMT cues were adjectives relating to nomothetic goal-statements. Results: Hierarchical multiple regression analyses found neither an overall effect of condition nor an interaction between condition and brooding on AM specificity. Thus, induced rumination over unresolved goals did not lead to higher levels of OGM than induced focus on resolved goals amongst individuals high in trait brooding. Multilevel hierarchical regression found that the extent to which people high on brooding were less specific in the unresolved condition did not depend on the importance or progress-discrepancy ratings of the goal-statements from which the MI-AMT cues were developed, nor on the relevance of the goal-cues to the concern identified in the goal cueing task. Goal-cue relevance ratings showed a significant main effect on AM specificity qualified by an interaction with condition whereby participants reported decreasingly specific AMs in response to cues related to the concern after the resolved goal manipulation. Conclusion: These null findings suggest that rumination over unresolved goals may not increase OGM amongst non-clinical samples. A replication of this study should utilise state rumination checks to ensure that the goal cueing task successfully differentially induced state rumination between conditions. Further exploration of the role of reflection might elucidate which qualities of rumination are positively associated with OGM but not present in rumination about unresolved goals. Given that Williams et al.’s CaR mechanism was constructed to understand OGM in clinical depression, a replication of this study using a clinical sample may be a useful next step in testing predictions made by this theory.

Objective: Moral reasoning (MR) reflects rationalisation in the moral domain, which matures across development and is underpinned by cognitive and affective processes. Although MR is associated with offending behaviours the mechanisms for this association are unknown. Examining the role of cognitive and affective processes in MR, and their influence on behaviour, may enhance existing psychological interventions that aim to reduce offending behaviours, and facilitate the development of novel targeted interventions. The current study investigated the hypothesis that MR would mediate the relationship between executive functions (EFs) and behaviour, and between empathy and behaviour. Method: In a cross-sectional design, typically developing adolescents (n = 72) individually completed an assessment battery, including the sociomoral reflection measure-short form, neuropsychological measures of working memory and cognitive flexibility/inhibition, and self-report questionnaires of empathy and behaviour. The battery also contained an assessment of intellectual functioning, and obtained data on socioeconomic status and age as confounding variables. Results: MR was not associated with self-report behaviour and, therefore, did not mediate the relationship between EFs/empathy and self-reported behaviour. A novel relationship was demonstrated between working memory and MR, and cognitive flexibility/inhibition was associated with MR. Self-report empathy was not associated with MR. Exploratory analyses suggested that intelligence and EFs were significant unique predictors of MR, and that truth and law moral values were associated with self-reported behavioural difficulties. Conclusions: Findings suggest that global MR is not associated with self-reported behaviour in typically developing adolescents, however, there may be an association between some moral values and self-reported behaviour. Findings also suggested that empathy is not associated with MR in this population, which warrants further investigation. These findings have implications for theoretical models of MR, and psychological intervention programmes. Recommendations for future research are presented.

Objective: To investigate whether traumatic brain injury (TBI) plays a role in impulsivity, risk-taking, aggression and other behaviour problems in young offenders. Setting: Youth Offending Teams in the South West of England. Participants: Participants were 21 young offenders with community sentences, aged between 14 and 18 years old. Design: A cross-sectional correlational design. Main measures: A background questionnaire gathered information about the frequency and severity of TBIs, which was used to calculate the “TBI dose”. The Stoplight task and the Stroop test provided behavioural measures of risk-taking and interference control respectively. Self-report questionnaires provided measures of impulsivity (UPPS Impulsivity Scale), aggression (Reactive and Proactive Aggression Questionnaire) and behaviour (Strength and Difficulties Questionnaire). Results: Seventy-six per cent of the sample reported at least one TBI. TBI was not significantly correlated with either any of the other variables. Impulsivity was significantly positively correlated with reactive aggression. Conclusion: This exploratory study concludes that, in the current sample, TBI is not associated with impulsivity or risk-taking. Therefore, it is equally possible that impulsivity and reactive aggression pre-exist TBI. Future research should use a longitudinal design and a larger sample.

Background: Self-help interventions can be useful in improving well-being and improving symptoms of low-level mental health problems. Goal-setting has also been found to be linked to well-being. In order to maximise the benefit of any goal related self-help interventions, it may be advantageous to consider for whom they may work best. Objective. The study aimed to replicate and extend previous findings relating to the efficacy of the Goal Setting and Planning (GAP) self-help intervention for improving well-being using an active control group to allow for a more carefully controlled test of GAP. In addition to this, the study aimed to examine the potential mediating effect that the intervention has via goal progress. Lastly the study aimed to investigate whether the gains in well-being associated with the GAP intervention will be greatest for individuals scoring more highly in ‘agentic’ value scores. Method. A controlled trial of 58 participants completed pre and post measures of well-being and were also led through a goal selection exercise in which they chose two goals to work towards over a five week period. Twenty-nine participants were allocated to the intervention condition which involved working through the GAP intervention in an online format to support their goal progress. The remaining 29 participants were allocated to the control group and were asked to work towards their goals without any further support. All participants were asked to provide goal attainment scores at the end of week three and at the end of week five. Results. There was no significant effect of condition on post-intervention outcome measures (p = .87, partial h2 = .02). Post measures indicated small but not significant gains in well-being across groups. Goal progress averaged across the two assessments did not differ significantly between groups. The conditions for mediation were not met therefore it was not possible to test for the mediating effects of goal progress on well-being. Lastly, moderation analysis showed that agency value scores did not significantly moderate the effectiveness of GAP in improving scores on any of the well-being measures. Conclusion: These null-findings particularly in relation to the efficacy of the GAP intervention were unexpected given significant findings in previous trials. However, the use of an active control group allowed for a more focused examination of the efficacy of the self-help components compared to previous trials. It is suggested that the goal-setting elements of GAP completed by both conditions may account for the small gains in well-being. Limitations of the current study are discussed particularly in relation to the use of a non-clinical, mainly undergraduate student sample. It is also suggested that goal-based self-help interventions may have limited use in non- clinical student populations. Ideas for future research are made including the recruitment of participants who would potentially benefit more from an intervention to improve well-being such as GAP.

Background: Exercise is recommended as a non-pharmacological intervention for individuals with a bipolar disorder diagnosis (BDD). Although physical activity can be beneficial for reducing depressive symptoms, there is preliminary evidence that high-intensity exercise can exacerbate (hypo)mania-related symptoms. Risks associated with other forms of exercise remain unknown. Method: To investigate the potential risks and benefits of low-intensity exercise, non-clinical participants were asked to either copy repetitive movements (n = 20), copy variable movements (n = 20) or watch variable movements (n = 21), following approach motivation induction. Hypomania-like symptoms, positive affect and approach motivation were measured pre-, during and post-task. Trait behavioural activation system (BAS) sensitivity was measured as a moderating factor. Results: There were no group differences in symptom change over time. BAS sensitivity did not moderate this relationship. Limitations: A predominantly student population with low average trait BAS sensitivity was studied. The reliability and validity of the approach motivation induction, mania measure and physical activity task are uncertain. Conclusions: It is unclear whether different types of low-intensity exercise are of risk or benefit for individuals prone to (hypo)mania. This area requires further investigation.

Background: Adults with autism experience high rates of social anxiety disorder (SAD). Cognitive models of suggest that anticipatory (AP) and post-event processing (PEP) develop and maintain SAD symptoms. These models underpin gold standard treatment. It is unclear if the cognitive model for SAD applies to an autism population. In addition, there is little understanding of how these symptoms are experienced by adults with autism. Aims: To examine whether the cognitive model of SAD is relevant for adults with autism, as indicated by correlations between SAD and cognitive maintenance factors. The second aim was to explore how adults with autism describe their experience of anxiety in social situations. Design: A mixed method approach was used. Participants (n = 30) completed questionnaires about characteristics of autism, SAD, AP and PEP. Bivariate correlations were used to investigate the relationship between cognitive processes and SAD. Partial correlations were used to explore relationships between SAD and cognitive processes while controlling for autism-specific language difficulties. Participants (n = 8) also completed a semi-structured interview about their social experience, associated anxiety and how any anxiety was managed. Thematic analysis was used to analyse responses. Results: Participants reported high levels of SAD, AP and PEP. There was no significant relationship between SAD and autism symptoms. While AP and autism were significantly positively correlated, exploratory analysis did not indicate a clear relationship between AP, PEP, SAD and autism. Results of thematic analysis indicate anxiety in social situations causes significant impairment, with distressing thoughts before and after social interaction. Conclusions: Results provide qualitative evidence that the Clark and Wells (1995) model of SAD is relevant for adults with autism. Further research is necessary to investigate the relationship between cognitive styles, such as intolerance of uncertainty, with SAD and autism. Adults with autism may benefit from CBT for SAD alongside strategies to structure interactions and the environment.

Literature Review: Background: Compassion-facilitating interventions (CFIs) seek to increase feelings of reassurance, safeness and well-being, and are suggested as a new psychotherapeutic approach to treat post-traumatic stress. Objectives: This review summarises and synthesises the literature investigating CFIs for post-traumatic stress disorder (PTSD) in clinical adult populations. Method: A systematic review of the experimental and intervention literature to date was completed using PubMed, PsycINFO, and Web of Knowledge databases. Results: CFIs show large to medium effect sizes in reducing PTSD in traumatised populations. However, many studies did not compare these interventions to active control conditions, and for the few studies that did, compassion-facilitating interventions were not better at reducing PTSD than the active control conditions. There was also not enough evidence in these studies to indicate a mechanism of change in these interventions. Conclusions: CFIs may be an effective intervention in PTSD, but more high quality research is needed to establish their efficacy over and above well established PTSD treatments. More research is also needed to identify the psychological mechanisms at work in decreasing PTSD symptoms Empirical Paper: Objective: Compassion-facilitating interventions are thought to be promising for treating post-traumatic stress disorder (PTSD). It is theorised that inducing self compassion through a compassion-facilitating meditation task would interrupt the psychological processes that cause emotional distress and trauma-related intrusions that form post-trauma. Methods: A student sample (n = 72) was exposed to a trauma-film and then either listened to a neutral-emotion audio mediation or a compassion-facilitating meditation (CFM). Self-reported distress, self-compassion and self-criticism were measured at baseline, post-film, post-audio meditation and for seven days following the experiment. Measures of sympathetic arousal and parasympathetic activation were measured at baseline and during the trauma film and audio meditation. Trauma-related intrusions were measured for seven days following the experiment. Results: Lower self-reported distress and higher self-compassion were reported over the seven day follow-up, but not directly after the meditation. No statistical differences between conditions over time were found for any of the physiological measures but skin conductance was higher in the CFM group, contrary to predictions. There were also no statistical differences found between the two conditions in terms of sum daily intrusions following trauma-exposure. Conclusions: This research supports the hypothesis that facilitating self compassion can disrupt some of the observed processes post-trauma, namely felt distress, although only after a delay. These findings are relevant to the theoretical models of PTSD and future clinical interventions.

Background: Motivational theories suggest that goals that are more internalised are associated with increased progress and well-being and that when goals are less internalised less progress is made. Rumination can occur as a result of unresolved goals and the nature of the goals that people set may be a factor in the degree to which they ruminate. Rumination has been associated with ill-being when it in unconstructive. Objective: The purpose of this study was to test whether different motives for goal pursuit predict unique variance in rumination, after accounting for other possible variables (goal importance, goal expectancy, goal conflict and goal facilitation) that may explain that relationship. Further investigation aimed to identify whether rumination about goals is perceived to be constructive, whether intrinsic and identified motives predict higher levels of goal progress and whether introjected motives and goal conflict each contribute significant variance in perceived constructiveness of rumination. Methods: This is a correlational diary study investigating goal rumination, constructiveness of goal rumination and goal progress on a daily basis. Forty-eight participants took part in the study (83.3% female [n=48]; age, M = 18.93 years, range = 18-43, SD = 7.72). Participants completed initial self-report measures on personal goals (Emmons, 1986), goal motivation (Ryan & Connell, 1989), goal importance and expectancy (Emmons, 1986), goal conflict and facilitation using unipolar scales (Riediger & Freund, 2004). They were then asked to complete a ten-day diary phase relating to their 6 most important goals. This included reporting on rumination (Schultheiss, Jones, Davis, & Kley, 2008), constructiveness of rumination and goal progress (Moberly & Dickson, 2016). Results: Support was found at the within-person level for introjected motives for goal pursuit being associated with higher levels of goal rumination and participants reported higher levels of rumination about goals that conflicted with other goals, both predicted unique variance. Introjected motives for goal pursuits were associated with higher levels of perceived constructiveness of rumination at the within-person level but not at the between-person level. Intrinsic and identified motives were not found to be associated with high levels of goal progress. Conclusion: This study was able to advance methodologically on previous studies finding that people ruminate more about goals pursued for introjected motives. Findings suggest introjected motives are associated with rumination and that this is not because of correlations with other variables such as conflict. This is support for the organismic integration understanding of rumination.

Objective: The current study aimed to explore associations between hypomanic personality traits, over-ambitious life goals, exercise motives, behavioural regulation and exercise behaviour. It was hypothesised that hypomanic traits would be significantly associated with setting more overly-ambitious life goals and extrinsic motives for exercise; these were derived from the theoretical perspectives of Self-Determination Theory (Deci & Ryan, 2000) and research into goal dysregulation in Bipolar Disorders (Johnson, 2005). Method: A sample of 165 undergraduate students (83% female [n = 133]; age, M = 19.3 years old, SD = 0.4, range 18-40) was recruited from the University of Exeter to take part in this longitudinal study. Participants completed a selection of self-report questionnaires measuring variables including hypomanic personality traits, over-ambitious goal setting, motives for exercise, behavioural regulation and exercise behaviour via an online survey. Exercise behaviour was then captured a week later in a follow-up survey. Results: Mediation analyses revealed an indirect effect of hypomanic traits on exercise motives via overly-ambitious goal setting. No significant relationships were found between hypomanic traits and exercise behaviour itself and further mediation analyses revealed no effect of exercise motives on exercise behaviour through behavioural regulations. Conclusions: Findings from this study provided mixed support for the proposed hypotheses; despite no significant associations between hypomanic traits and exercise behaviour itself, hypomanic traits appear to make a unique contribution to exercise motives, via over-ambitious goal setting. Therefore, they may play a role in the motivational processes that have consistently shown to be associated with exercise engagement and adherence (Ingledew & Markland, 2008, Teixeira et al., 2012). Exploring these variables within a clinical population may be a useful next step in further understanding these motivational processes.

Objective: Neuropsychology models of anorexia nervosa (AN) propose that cognitive difficulties including poor Emotion Recognition (ER) and set-shifting ability may be central to the development and maintenance of eating pathology. This study aimed to test the central positions of such models by assessing specific ER difficulties in AN as well as the relationship between ER deficits and set-shifting performance. Methods: Fifty-one women were assessed (25 with AN; M = 28.20 SD = 8.69 and 26 control M = 21.27 SD = 5.10) on a novel measure of ER, a set-shifting test and self-report questionnaires concerning co-morbid factors. Results: The data did not reveal a global difference in ER or set-shifting performance between groups. Specific hypotheses of ER deficits in AN were also not met as performance on individual emotions was comparable between groups. There was an unexpected negative correlation between disgust recognition and set-shifting performance, however, this was only significant across the whole sample. ER performance was not related with any confounding factors. Conclusions: Despite an abundance of research supporting the position of social cognitive difficulties in AN, the current study failed to find global or specific deficits in ER in the present sample. Similarly, ER performance was not related to set-shifting as proposed by neuropsychological models of AN aetiology. Possible explanations for a lack of difference observed using this novel ER task are explored and future directions for evaluating ER in AN are discussed.