Background: Socio demographic and epidemiological change, together with the availability of medical technology, may prolong the process of dying. Antecedent control over end-of-life care (EoLC) may be sought by making an advance directive (AD). This option has been legalised in some countries including Israel, the setting of this study. In Israel, EoLC is set in the context of complex and interrelated religious and state regulations. Aim: To examine from a variety of perspectives, the experiences, beliefs and practices associated with the use of ADs in Israel, with a view to understand their role in the lives of patients with palliative care needs towards the EoL. Methods: The study was conducted in 2011, using mixed methods in two phases. The first phase involved qualitative methods to construct patient-centred case studies. Twenty seven participants (patients, relatives, physicians and a Rabbi) were interviewed. In phase 2, a survey examined health care providers’ (HCP) knowledge, attitudes and experiences regarding ADs. The final sample (n=72) of HCPs comprised: 57 nurses (79%), 8 physicians (11%), and 7 social workers (10%). Findings: This study promotes the understanding that ADs reflect a multitude of steps that are distinct but linked, as in a ‘relay’. The findings demonstrate complexity: patients’ ADs reflected their desire for autonomy but their decisions were often influenced by their relationships. In turn, relatives feared making mistakes in decision-making, while also feeling obligated to enact the patient’s wishes. Physicians reported finding it difficult to forgo active treatments and found communication with relatives about patients’ ADs challenging. The survey showed that while HCPs had typically positive attitudes towards, and some experience with ADs, they also perceived some barriers towards their use. Conclusion: ADs seem as an imperfect solution to the complexities surrounding EoLC planning, and additional solutions than those available today may be required.

In 2004 unexplained pedagogical barriers were limiting Learning Object (LO) development. Few reference points existed preventing the formation of specific pedagogical questions as to the nature of these barriers - hence this PhD's rationale. This thesis 'uncovers' the most effective pedagogical and evaluation/assessment* approaches for LO design in Medical Education, and the underlying principles within these approaches - i.e. what is effective, and why. To determine why certain approaches are effective observation/interview/usability studies were performed using grounded theory to generate hypotheses (1A Participants n=57). To verify 1A findings, this process was replicated using different sites/samples in Phase 2(Eastern/Midlands, n=72). To determine what was most effective, systematic reviews using a purpose-built design were undertaken with additional questions on pedagogy and evaluation/assessment• components (1B Studies n=222). Approaches identified as 'effective' according to statistics, SCIE and my own rigor scoring systems were tested blind in two locations (Eastern/Midlands) with different samples under a null hypothesis (i.e. 'Each approach will score no differently to any other', Phase 2 participants n=72). This was further developed by replicating this process via mobile delivery. Section 1A generated over a hundred hypotheses. In Section 1B, two existing approaches scored consistently high. Phase 2 produced the same hypotheses/approaches when submitted to the blinded observation/interview/usability process thus tight theme linkage resulted in rigorous theory and empirical data. The two top-performing 1B approaches scored high resulting in the possible existence of generic principles. When replicating 1A, 1B and Phase 2 for mobile delivery, the existence of generic principles was verified and a possible model for practice formed. In summary, this thesis underlines the importance of learner input and how learners' perceptions form an essential part of the LO learning process. It discovers original generic principles for both desktop and mobile formats, highlights how branch and loop learning systems are necessary for learner customisation, and provides new knowledge verifying Wiley's molecular LO analogy. *In this thesis many types of evaluation approaches are tested. These are called 'evaluation approaches' by the authors that created them. However, in some disciplines the term 'evaluation' is viewed as being interchangeable with the term 'assessment'. For this reason explanatory footnotes will be given throughout where necessary.

Background: The Quality Outcomes Framework (QOF) was introduced in April 2004. Research to date has primarily looked at its effects on process and surrogate outcomes measured in primary care rather than evidence linked hospital admissions. None has used individual linked data. This study utilises data linkages between the Clinical Practice Research Datalink (CPRD) and Hospital Episode Statistics (HES) datasets to determine the impact of QOF targets on hospital admissions at the individual patient level. Methods: CHD QOF targets and linked hospital admissions were selected according to the strength of their evidence base and ease of extraction. Outcomes were ICD10 codes for Ischaemic Heart Disease (IHD). These formed the primary diagnosis over a hospitalisation and depending on the severity of the code, were additionally an emergency admission. Econometric analysis was then undertaken with IHD admissions as the dependent variable and evidence based CHD QOF targets as explanatory variables. Results: Evidence based CHD QOF targets were found to significantly reduce outcomes after a one year lag. Of the co-morbidities included, only Heart Failure was consistently found to significantly increase outcomes in all analyses. Higher deprivation, and having a study outcome prior to CHD diagnosis, significantly increased outcomes. Being treated in a higher performing practice on the selected targets, in itself, significantly reduced outcomes. Conclusion: This study has demonstrated at the individual level that evidence linked targets in the QOF are effective in reducing linked hospital admissions with a lag. It is the first study to take advantage of CPRD and HES linkages to do so. This has been demonstrated in a ‘real’ world setting, outside of controlled clinical settings, and in so doing addressed deficiencies identified in the existing research. This research has shown that large administrative datasets can support such research and opened up a number of possibilities for future research.

Background Effective disease diagnosis and treatment relies on a conceptual knowledge base that is both expansive and well-networked. The problem-based learning (PBL) curriculum is considered as being well-suited to creating this kind of knowledge. The facilitator plays a crucial role in establishing and maintaining the knowledge construction discourse as students interact to resolve case problems. An exploration of tutorial talk could provide opportunities to understand and improve verbal interactions of this nature. Many of the previous studies have only analysed a small amount of tutorial talks owing to methodological constraints, and the existing literature on the subject matter only scarcely touches upon the utility of lexicogrammatical methods for the development of an understanding of knowledge construction in medical PBL tutorials. In this research, a blend of corpus linguistics methodology and a lexicogrammatical approach was employed for the analysis of talk in 8 PBL tutorial groups in order to deepen our understanding of how students jointly construct knowledge and how the facilitator guides the process. Aims In this study, a corpus of 2,37,820 comprising eight PBL students’ and facilitators’ tutorial talk was created to achieve the following aims: I. To use the students’ subcorpus to answer the research question (1) by measuring the frequencies and describing the functions of the frequently occurring (1) referring expression indicators; (2) shared knowledge indicators; (3) knowledge extension indicators; and (4) knowledge enhancement indicators. II. To use the facilitators’ subcorpus to answer the research question (2) by measuring the frequencies and describing the functions of the commonly occurring (1) facilitators’ questions; (2) facilitators’ directive expression indicators; and (3) facilitators’ probability indicators. III: To make recommendations based on the results of the study. Methodology Wmatrix 3 was used to retrieve defined linguistic indicators relating to the research questions. A quantitative analysis of the indicators was performed through word frequency computation and a keyword-in-context analysis. Descriptive statistics with SPSS version 22 was used to computer frequency profile of the indicator functions, and the Log likelihood calculator was used to determine the variation of the functions across the eight PBL groups. Extracts from the dataset were provided to illustrate the indicators’ functions. I. Results of Students’ talk analysis The subcorpus contained 2,10,077 words. The most frequent contents of the students’ talk comprised biomedical science and cause-effect vocabularies. 1. Analysis of referring indicators There were 2,325 referring expression indicators. They were used to mark verbal expressions, amounting to 44.04%; mental expressions, amounting to 42.24%; and learning situation and materials, amounting to 13.72%. The referring expressions were used for providing peer commendation, sharing knowledge, fostering social and cognitive regulation, and for constructing knowledge; the mental referring expressions were used to generate hypotheses, achieve mutual understanding, and define group tasks; and learning referring expressions were used to share learning resources, explain concepts, as well as guide discussions and resolve conflicts. 2. Analysis of shared knowledge indicators There were 3,437 shared knowledge expression indicators, which are the following: affirmation (73%), negation (17%), and non-lexical content (10%). Affirmative indicators were mostly used for integration-oriented knowledge sharing (42.31%); negation affirmation expressions were mostly used for conflict-oriented knowledge sharing (70%); and non-content indicators were mainly used for idea and information orientation. Shared knowledge was commonly achieved among group members through information addition, repetition and rephrasing, paraphrasing, causal and noncausal elaboration, correction of ideas and information recollection, and by establishing orientation to ideas and information from the group members. 3. Analysis of knowledge extension indicators There were 6,520 retrieved knowledge extension indicators, which comprised the following: additive 4,227 (63.54%), alternative 1,001 (15.05%), and adversative 1,424 (21.41%). Adversative indicators were more frequently used for knowledge construction compared to additive (33% versus 16%; LL 32.58, p < 0.01) and alternative indicators (33% versus 13%; LL 95.74, p < 0.01). The students commonly used additive indicators for simple, temporal, causal-conditional, elaborate, contrastive, and indefinite additions. Alternative indicators were commonly used for offering alternative questions and ideas while adversative indicators were frequently used to link elaborative, contrastive, concessional, and causal-conditional clauses. 4. Analysis of knowledge enhancement indicators A total of 6,402 indicators were retrieved. The most frequent among the retrieved 6,402 indicators were because, so, as, when, and that. Between 16.94% and 29.24% of the indicators were used for knowledge co-construction. The most frequent indicators’ functions were conditional, extension, report, consequence, inference, and feature specification. The reporting functions regularly concerned biomedical theory, previous peer knowledge, research evidence, professional opinion, as well as cognitive tools and criticism; extension function related to biomedical knowledge; and feature specification functions involved biomedical attributes and explanation; the conditional functions were frequently used to state logical conditions for disease presence, manifestation, and treatments; the inferential functions were more consistently used to link biomedical deductions to their premise; and the consequential functions commonly related to the linking of physiological mechanism and organ function to their respective consequences. II: Results of facilitators’ talk analysis The subcorpus contained 27,743 words. The most frequent content comprised biomedical science and cause-effect vocabularies. 1. Facilitators’ questions There were 35 types of question indicators. The facilitators asked 0.78 lower-order questions per 100 tokens, and 0.25 higher-order questions problem-based per 100 tokens. The questions functioned to stimulate elaboration, elicit information, prompt students, and the offering of suggestions. 2. Analysis of directive expressions ‘Should’, ‘have to’, ‘need’, ‘supposed’, ‘would’, and ‘can’ directive expression indicators were found to be most frequent. They were used to mark expectation, indirect question, and they were commonly used to preface requirement, exhortation, and intention. The indicators functioned frequently to facilitate group process (53.45%) and direct learning (42.00%). 3. Analysis of probability expression indicators There were of 27 types, out of which 9 were frequently occurring. The probability expression indicators were frequently used to mark possibility, prediction, hedging, and logical deductions. The indicators functioned to preface content information given to the students, to mark process facilitation remarks, and to mark facilitators’ questions. Discussion and conclusion The study demonstrated the feasibility of using corpus linguistics to study medical students’ knowledge construction talk; provided evidence of knowledge construction through prior knowledge mobilisation, knowledge extension, and enhancement; and signified the attainment of shared knowledge. The facilitators frequently asked lower-order questions; the directive expressions indicators were used to mark content-related and learning behaviour expectations and requirements; and the probability expression indicators were frequently used to mark content information given to the students. This study shows that students construct knowledge in their PBL tutorials. The pedagogic issues that emerged from the study relates to subversion of the PBL facilitation principles. A wholistic understanding of the factors that affect the behaviours of the facilitators in the classroom is important to resolve this problem. This may involve tutor pedagogic education and recalibration of administrative policies and institutional culture to provide an enabling environment for PBL instructional approach.

This thesis explores and tests an integrated, social cognitive approach to understanding leadership and developing leaders within healthcare organisations in the United Kingdom. Two studies were implemented; the first involved an assessment of the personal and environmental components of leaders and their followers, as well as the self-reported and follower-rated behaviour of leaders in eight National Health Service (NHS) organisations. The second study involved working with 15 organisations across the NHS to implement and evaluate a scenario-based training intervention which promotes a social cognitive perspective of leadership and draws on the principles of adult learning. The impact of this intervention was examined using a longitudinal design and factors influencing the transfer of leader self-efficacy were explored. For both studies, the suitability of measures was assessed and validated using confirmatory factor analysis. Specific analytic techniques were used within each study based on the study objectives. In study I, mediation analyses were conducted using Structural Equation Modelling (SEM) to explore the interaction between personal, environmental and behavioural components within leaders, as well as exploring how these components interact in the relationship between the leader and their followers. The results revealed a significant mediational role of two key aspects of the perceived environment (control and social resources) in the relationship between leaders’ behaviour and followers’ personal components and attitudes towards work. The mediational role of leader behaviour in the relationship between leaders’ personal components and their perceived environment was not supported. Study II used a Multivariate Analysis of Variance (MANOVA) with a repeated measures design to explore the impact of training over time on leaders’ awareness of what constitutes ‘supportive’ leadership behaviour and their self-efficacy judgements. Data was collected immediately prior to attendance at the training intervention, immediately after training and 6 months post-training. The results revealed a significant increase in leaders’ self-efficacy post-training in relation to four common situations that leaders within healthcare settings face. The longitudinal analysis also showed that leaders’ self-efficacy beliefs were maintained 6 months post-training. In addition, an assessment of leaders’ behavioural awareness indicated that democratic leadership behaviours and behaviours relating to greater self/other awareness were seen to be significantly more supportive immediately after training compared to pre-training (baseline) levels. There were no significant differences in leaders’ awareness of the supportiveness of performance and target-driven behaviours immediately after the training. However, the longitudinal analysis revealed that leaders’ increased awareness of the supportiveness of these behaviours was not sustained over time. Instead, these behaviours were seen to be significantly less supportive 6 months after training compared to both the immediate post-training and pre-training levels of awareness. Finally, a Hierarchical Multiple Linear Regression (HMLR) analysis revealed that the transfer of leaders’ self-efficacy beliefs (as assessed 6 months post-training) was significantly predicted by leaders’ pre-training (baseline) and post-training self-efficacy levels, as well as leaders’ age. Contrary to what was expected, application of their learning, post-training support from colleagues and supervisors and the interaction between support and the application of learning did not predict leaders’ self-efficacy 6 months after training. This thesis concludes by discussing the implications for leadership research and leadership training within healthcare organisations.

The aim of this thesis is to examine the impact of polygraphy disclosure and recidivism (particularly for sex offenders), gain greater understanding of some of the psychological mechanisms involved, and, in the light of this, consider the value of its application in forensic settings. Chapter one presents a general introduction to the topic. Chapter two presents a systematic review of the relevant literature. In particular, it explores the utility of postconviction polygraph testing amongst sexual offenders, with a primary focus upon its influence in facilitating disclosure. The findings are generally supportive of the view that the polygraph is a useful technique for eliciting additional information from offenders, which, in turn, can assist in achieving a more accurate understanding of current risk and criminal history. It is noted, however, that there continues to be a paucity of high quality research evidence, in particular, a lack of adequately controlled studies. Further research is needed to gain more secure understandings of the polygraph’s potential. Chapter three offers a critique of the use of the postconviction polygraph in the monitoring, assessment and supervision of convicted sex offenders. This notes a degree of partisanship in arguments for and against its use. The chapter seeks to offer an analysis of available evidence concerning the utility of the post-conviction polygraph in encouraging disclosures amongst sexual offenders under investigation in both custodial and noncustodial settings. Those who support the use of the polygraph typically argue that such disclosure provides fuller histories of deviant sexual behaviour, admissions of previously unknown offences/victims, and increased reporting of other high-risk behaviours. However, noting the concerns of those opposed to its use, the ongoing challenges and shortfalls of the polygraph are also referenced. In the light of the analysis, it is proposed that the method should not be used in isolation but, where appropriate, in conjunction with a range of other assessment tools before reaching a risk-related conclusion concerning sexual offenders. The ‘bogus pipeline to the truth’ effect is an argument used to challenge the veracity of polygraph data, as its underlying premise is that the value of the polygraph in eliciting disclosure is not grounded in its accuracy, but, rather, in false claims regarding its ability to detect deception (which, as a result, will often elicit more truthful responses from those being tested). In other words, disclosure can be derived from psychologically manipulating examinees by convincing them that the lie detector is significantly more accurate than it actually is. In chapter four, an empirical research study explores the bogus pipeline effect upon disclosures made by students who were earlier exposed to cheating behaviour. Participants experienced within-group cheating by a study confederate and later placed in one of three conditions in which they were asked about the occurrence of cheating within their group. Two groups were attached to a bogus lie detector and informed that this was either 75% accurate or 100% accurate. Control participants were not attached to the lie detector, but simply asked if they had been present in a group that had cheated. Findings demonstrated that those participants attached to a bogus lie detector were more likely to disclose cheating within their group than those asked in the absence of the machine. However, there were no significant differences in rates of disclosure between the 75% and 100% condition. The influence of suggestibility, personality and gender upon disclosure amongst all conditions revealed no effects confounding the relationship. A single case study is presented in Chapter five. It details some of the ways by which the psychologist can encourage a client with paranoid schizophrenia, and a history of sex offending, to reflect openly upon his psychological and behavioural circumstances. Subsequently, it considers whether a measure such as the polygraph, designed to encourage honest accounts and attributions, could be of value in work with those clients whose thought processes are distorted by psychotic conditions. In contextualising this issue, the chapter describes and evaluates the usefulness of a ten week psychosis awareness group programme for a sexual offender treated within a medium security hospital. The influence of this psychoeducation programme on levels of understanding, disclosure and personal acceptance of his mental illness is discussed, and recommendations for further psychological work are made. Subsequently, the association between psychotic symptoms and sexual offending is explored, as is the suitability of a future polygraph for this client In Chapter six a synoptic discussion of the work presented concludes the thesis. This addresses the current position on the bogus pipeline paradigm. Finally, the chapter seeks to provide an informed position concerning the use of the post-conviction polygraph in forensic settings.

BACKGROUND: Problem Based Learning (PBL) facilitators have sometimes sought to motivate learners to ‘talk more’ in order to ‘learn more’ but there has been little work to differentiate through differences evident in their speech behaviour during PBL discussions and to identify which differences are associated with successful learning. Corpus analysis is a useful tool for probing students’ speech in PBL, providing a quantitative basis for examining student talk. AIMS AND OBJECTIVES: The aim of this study was to create a speech corpus of at least 100,000 words, which identifies each individual student’s contribution to a PBL discussion of the same case at the same time within six randomly selected, gender-balanced PBL groups. The study then uses corpus analysis to explore the nature of the speech of individuals and their performance in learning. A statistical analysis was designed to triangulate findings through measuring and analysing underlying factors thought to influence the speech behaviour of individual learners, such as gender and personality type and to identify correlations with educational attainment. METHODS: Wmatrix3 was used to identify significant differences in frequency, parts of speech and semantic category usage from verbatim transcripts of the same three session cycle of PBL. SPSS analyses identified significant differences between learners’ speech behaviour, personality type, gender and attainment in the first year. RESULTS: Multiple and simple linear regression analysis demonstrates significant correlation to a moderate degree between certain data points. In summary, the calmer, more conscientious, lower openness to experience learner, with a high GAMSAT score, achieved significantly higher MCQ scores at the end of year one. Introverted students spoke less frequently, but with longer utterances. Male students spoke significantly more frequently than female students. More emotionally reactive (neurotic) students achieved significantly lower scores on knowledge tests at the close of year one and their talk focussed their own performance. Corpus analysis demonstrated statistically significant differences between personality, gender, high and low test performers. Lower performers frequently dwelt on ‘what’ and ‘why’ and more general technical terminology. Higher performers used ‘when’ and ‘how’ and highly specific technical vocabulary more frequently. CONCLUSION: This study demonstrates that corpus analysis of PBL transcripts provides rich evidence of significant differences in individual speech behaviour whilst learning through discussion. The results broadly replicate well-established findings about gender communication styles and personality difference. This suggests that there may be a utilitarian case for enhancing the approach of PBL facilitators and learners to individual differences without resorting to the overgeneralisations of stereotyping. For example, male and female learners often demonstrate different communication strategies, which may potentially limit the breadth of their future clinical communication skills. Facilitators and learners may use this awareness to explicitly encourage both male and female learners to adopt the full range of communication behaviours. It is important for facilitators to model best practice in their own communication in the initial sessions. By making such language patterns evident, it becomes possible to ensure that PBL groups become more skilled in initiating spontaneous contributions from introvert and female learners, in encouraging longer contributions from extraverts, more factual statements from female students and more varied types of speech from male learners. Learners who identify themselves as emotionally reactive (neurotic or volatile) could productively seek to moderate unproductive anxiety from an early stage through adopting excellent study habits and attitudes, particularly with respect to assessments.

Background For epidemiological research purposes structured data provide identifiable and immediate access to the information that has been recorded, however, many quantitative recordings in electronic medical records are unstructured. This means researchers have to manually identify and extract information of interest. This is costly in terms of time and money and with access to larger amounts of electronically stored data this approach is becoming increasingly impractical. Method Two programmatic methods were developed to extract and classify numeric quantities and identify attributes from unstructured dosage instructions and clinical comments from The Health Improvement Network (THIN) database. Both methods are based on frequently occurring patterns of recording from which models were formed. Dosage instructions: Automated coding was achieved through the interpretation of a representative set of language phrases with identifiable traits. The dosage data table was automatically recoded and assessed for accuracy and coverage of a daily dosage value, then assessed in the context of 146 commonly prescribed medications. Clinical comments: Automated coding was achieved through the identification of a representative set of text and/or Read code qualifications. The model was initially trained on THIN data for a wide range of numeric health indicators, then tested for generalizability using comments from an alternative source and assessed for accuracy, sensitivity, and specificity using a subset of 12 commonly recorded health indicators. Results Dosage instructions: The coverage of a daily dosage value within the dosage data table was increased from 42.1% to 84.8% coverage with an accuracy of 84.6%. For the 146 medications assessed, on a per-unique-instruction basis, the coverage was 79.7% on average with an accuracy of 95.4%. On an all-recorded-instructions basis the weighted coverage was 65.9% on average with an accuracy of 99.3%. Clinical comments: For all 12 of the health indicators assessed the automated extraction achieved a specificity of >98% and an accuracy of >99%. The sensitivity was >96% for 8 of the indicators and between 52-88% for the other indicators. Conclusion Dosage instructions: The automated coding has improved the quantitative and qualitative summary for dosage instructions within THIN resulting in a substantial increase in the quantity of data available for pharmaco-epidemiological research. Clinical comments: The sensitivity of the extraction method is dependent on the consistency of recording patterns, which in turn was dependent on the ability to identify the differing patterns of qualification during training.

Thesis Aim: Boundary violations are committed when a professional crosses the emotional, physical, spiritual or sexual limits of another, thus disrupting the relationship between staff and clients. This thesis explores this issue in the context of healthcare and forensic settings. It identifies boundary violation risk and protective factors for staff working with forensic healthcare patients. The thesis also investigates reasons why staff under-report and may not whistleblow their concerns in these areas. Methods: A mixed method approach was used to explore multiple elements of this topic. A systematic review was used to identify risk factors associated with staff committing boundary violations in healthcare and forensic settings. This was then expanded, through a regression, to specifically look at risk and protective factors in the forensic healthcare settings. A thematic analysis was carried out to explore staff’s perceptions of whistleblowing and boundary violating behaviours. A case study and several psychometrics (including the Exploitation Index) were also used to investigate this topic in more depth and assess the impact that boundary violating can have on the patients involved. Findings: The setting and environment were shown, through the systematic review, to be key when exploring this topic, as they not only impact on the definition of boundary violating along with what factors protect or put staff at risk of boundary violating. The regression identified risk factors, including staff feeling they had no one to turn to and working alone with patients. There were also factors which protected staff from committing boundary violations, such as having good social support and receiving training about boundaries. The thematic analysis found that people did not raise concerns of colleagues’ boundary violating due to concerns about anonymity, feeling that nothing would be done, and feeling that personally intervening (not management) would be a more effective method. Conclusions: It is of great importance to explore this topic and develop a greater understanding of boundary violations, as the impact of and harm caused by this behaviour are wide ranging. This thesis can aid the identification of professionals at risk, and create strategies to prevent or manage this destructive behaviour. These findings can also help create a more open and honest arena for concerns about boundaries to be raised and managed.

The majority of research on academic procrastination has been conducted among undergraduate students, and there is relatively little research investigating procrastination among doctoral students. PhD students are different from undergraduates: they might need a higher level of self-regulatory ability to conduct research. The aim of this thesis was to fill the research gap in the investigation of academic procrastination among PhD students in the UK. More specifically, the current investigation combined different perspectives to examine the extent to which PhD students procrastinate, explores the relationships between a variety of psychological variables, doctoral satisfaction and academic procrastination, and identifies the antecedents and influence of procrastination in relation to PhD students’ own experience. This thesis comprises three studies. Firstly, a cross-sectional study (N=285) was conducted in order to assess the relationship between doctoral satisfaction, Big Five personality traits, self-efficacy, self-esteem, anxiety, depression, and academic procrastination. In Study 2, a longitudinal research design was employed to examine the stability of the identified relationships over a 12-month period (N=79). The results indicated that doctoral satisfaction and depression had long-term influence on PhD students’ procrastinatory behaviour. In addition, conscientiousness was found to have an effect on academic procrastination only in the cross-sectional study, while openness was found to predict procrastination longitudinally. Moreover, doctoral satisfaction also had meditional effect on the relationship between personality traits, self-efficacy, anxiety, and procrastination. In Study 3, the antecedents and consequences of academic procrastination in terms of UK PhD students’ own perspectives were explored in twenty-one in-depth interviews. Data were thematically analysed and a description of the themes concerning antecedents, positive and negative consequences of procrastination, and coping strategies used to help reduce procrastination, is provided. The findings indicates that PhD students’ procrastination is a multifaceted phenomenon with cognitive, affective, and behavioul factors influencing its likelihood. Causes and effects of academic procrastination among PhD students are discussed on the basis of findings from the quantitative and qualitative studies, by considering individual differences, psychological state, and contextual factors in a new conceptual model of academic procrastination. The findings point to a range of possible procrastination-reduction interventions focused on doctoral satisfaction and psychological wellbeing. The strengths and limitations of this work are discussed.