Search

Recommender systems have become ubiquitous in the past years. They solve the tyranny of choice problem faced by many users, and are utilized by many online businesses to drive engagement and sales. Besides other criticisms, like creating filter bubbles within social networks, recommender systems are often reproved for collecting considerable amounts of personal data. However, to personalize recommendations, personal information is fundamentally required. A recent distributed learning scheme called federated learning has made it possible to learn from personal user data without its central collection. Consequently, we present a recommender system for movie recommendations, which provides privacy and thus trustworthiness on multiple levels: First and foremost, it is trained using federated learning and thus, by its very nature, privacy-preserving, while still enabling users to benefit from global insights. Furthermore, a novel federated learning scheme, called FedQ, is employed, which not only addresses the problem of non-i.i.d.-ness and small local datasets, but also prevents input data reconstruction attacks by aggregating client updates early. Finally, to reduce the communication overhead, compression is applied, which significantly compresses the exchanged neural network parametrizations to a fraction of their original size. We conjecture that this may also improve data privacy through its lossy quantization stage., Comment: Accepted for publication in the ACM Transactions on Recommender Systems (TORS) Special Issue on Trustworthy Recommender Systems

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), Queensland Conservatorium, Arts, Education and Law, This study investigated the effectiveness of flow teaching as directed to the practice and performance of instrumental music in the post-secondary context. Its purpose was to determine the effectiveness of a three-lesson educational flow intervention for twelve post-secondary brass and string instrumentalists. Data were collected through measures of subjective experience during practice and performance and the objective assessment of three performances. Flow experiences have been studied extensively in sport since the 1970s, and consequently, training strategies have been developed to increase flow levels in sporting contexts to maximise the possibility of optimal experience. However, in the domain of instrumental music, flow training strategies have not been thoroughly developed, researched, or applied for practice or performance purposes. Therefore, it is not known what potential effects, if any, applied flow theory strategies might have on instrumental practice and performance experiences or performance outcomes. The study employed a mixed method approach using recognised qualitative and quantitative methods suitable for an educational intervention research investigation. The methodology was a quasi-experimental intervention with non-equivalent control using concurrent triangulation comparative mixed method design. This provided the simplest form of baseline design with two groups and allowed each group to act as both an intervention group and a control group maximising the possible data collection for the small sample and short timeframe. The participants in the study were 12 self-selected undergraduate string and brass players from the Bachelor of Music course at the Queensland Conservatorium Griffith University (QCGU), Australia. They were stratified and randomly assigned into two groups of six. During the Intervention phases of the 38-day study, each participant received three one-to-one Flow Music Method (FMM) lessons created by the researcher. The FMM was b

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Applied Psychology, Griffith Health, The aim of the current research program was to empirically investigate the experience of peak performance in the context of professional screen acting. Five studies were conducted employing an exploratory sequential mixed methods research approach. Three qualitative studies explored the subjective peak screen acting performance experience. Semi-structured interviews were conducted with five professional screen actors. The first two studies employed an interpretative phenomenological analysis. The first study revealed that peak screen acting performances were a present, connected, and heightened state of living as the character (i.e., the peak performance state). Peak screen acting performances were achieved through performance preparation. Actors reported developing self-efficacy and acceptance due to feeling prepared. The second study revealed that professional interpersonal relationships either facilitated or inhibited screen acting performances. Further, intrinsic motivation facilitated, whilst performance pressure inhibited peak performances. The third study employed a deductive frequency analysis to explore whether flow was related to peak screen acting performances. Flow was partially associated with peak screen acting performances. Experiential dimensions of flow were most strongly associated with peak performances. The combined results of the first three qualitative studies indicated that a three-level hierarchy of needs may best represent the process that assists actors reach peak performance in screen acting. The final two studies used quantitative methodology to extend upon the first three studies. Participants were professional actors with various levels of professional experience. Participants completed a self-report online survey measuring general performance, the peak performance state, performance preparation, support, trust, respect, self-efficacy, psychological safety, acceptance, intrinsic motivation, pressure,worry, and flow. The fourth study (N

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Applied Psychology, Griffith Health, Theory of Mind (ToM) is the ability to attribute mental states, including knowledge, beliefs, and intentions to oneself and to others (Premack & Woodruff, 1978). It is crucial to investigate ToM because of its central importance in everyday social communication and interactions (Baron-Cohen, 1999). Most previous studies have focused on ToM in pre-schoolers (Liu et al., 2008; Wellman et al., 2001). Although ToM continues to develop into middle childhood (Miller, 2009), little is known about school-aged children’s ToM across cultures. Therefore, the present study aimed to clarify age-related differences of school-aged children’s ToM and its sociocultural correlates and underlying cognitive mechanisms in two cultural samples of 5.5 - to 12-year-old children from China (n = 126) and Australia (n = 83). The overall project has been broken into three studies. Study 1 assessed Chinese and Australian children’s ToM using four measures, which reflect three categories: cognitive ToM (Interpretive ToM [IToM]), affective ToM (Faux Pas and Reading Mind in the Eyes [RMIE]), and mixed task (Strange Stories). The results revealed that younger children aged from 5.5 to 7 years in both cultural groups performed worse than older children aged from 8 to 11 years on all measures. Australian children outperformed Chinese children on both RMIE and Strange Stories. In contrast, no cultural differences were found in children’s scores on IToM and Faux Pas. However, an analysis of types of errors on the two tasks reflected cultural influences. Chinese children made more reality/contamination errors in the IToM task than Australian children, whereas 10- to 12-year-old Australian children made more identical responses than their Chinese age cohorts. Moreover, Australian children made fewer errors in Faux Pas identification than Chinese children. Study 2 examined how parenting practices and sibling composition influenced ToM across cultures. The results revealed that corporal punishment negative

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Applied Psychology, Griffith Health, Examining methodological differences in fear conditioning and extinction studies is a growing area of research, inspired by the recent ‘replication crisis’ in psychological research. Determining whether methodological differences of fear conditioning and extinction paradigms contribute to inconsistent findings is important in terms of the translational value of this paradigm as an experimental analogue of exposure-based cognitive behavioural therapy. The aim of this thesis was to improve the experimental methodology of fear conditioning and extinction studies, by addressing three key questions: 1) Are there parameters of fear conditioning and extinction tasks that differentiate successful from unsuccessful extinction in studies with children and youth; 2) Do different types and combinations of within-phase subjective measures affect within-phase physiological and between-phase measures of fear conditioning and extinction; and 3) Does behavioural memory reconsolidation in the form of cue reactivation with versus without instructions differentially impact extinction and return of fear? Question one was addressed by conducting a systematic review of 35 fear conditioning and extinction studies conducted with children and adolescents (Chapter 2, Ryan et al., 2019). This review revealed that successful extinction was observed in studies that used fear irrelevant shapes (CS) and tones (US) (recommended for all youth) or fear relevant faces (CS) and screams (US) (recommended only for adolescents) as well as those studies that used 8 – 12 CS trials per acquisition and extinction phase. The most commonly used and effective dependent measures were skin conductance responses (SCRs) and subjective ratings of CS valence, fearfulness and arousal. Fear potentiated startle (FPS) was also effective, however less commonly used with children. It was suggested that standard measures be adopted for subjective ratings scales to assist in comparisons across studies. It was recommended that

In large-scale wireless networks, effective transceiver design based on imperfect knowledge of the wireless channel is crucial. In this work, we introduce methods for transceiver design and resource allocation that make use of asymptotic results to reduce computational complexity. We focus on linear transceivers, which are close to optimal for the system dimensions we have in mind. Since the introduced methods rely heavily on second order statistics of the wireless channel, we further discuss efficient methods for the acquisition of these statistics in practical systems., Die Entwicklung von effektiven Algorithmen für Sende- und Empfangsverarbeitung auf Grundlage von imperfekter Kanalinformation spiel eine Schlüsselrolle in hochdimensionalen drahtlosen Kommunikationssystemen. In dieser Arbeit stellen wir entsprechende Algorithmen vor, welche asymptotische Ergebnisse ausnutzen um die Komplexität zu verringern. Wir setzen den Fokus auf lineare Sende- und Emfpangsfilter, da diese für die hochdimensionalen Systeme die wir betrachten nahezu optimal sind. Da wir für unser Filterdesign auf die Verfügbarkeit von statistischer Information über den Kanal angewiesen sind, diskutieren wir des weiteren effiziente Methoden um diese statistische Information zu schätzen.

Theory of Mind (ToM) is the ability to attribute mental states, including knowledge, beliefs, and intentions to oneself and to others (Premack & Woodruff, 1978). It is crucial to investigate ToM because of its central importance in everyday social communication and interactions (Baron-Cohen, 1999). Most previous studies have focused on ToM in pre-schoolers (Liu et al., 2008; Wellman et al., 2001). Although ToM continues to develop into middle childhood (Miller, 2009), little is known about school-aged children’s ToM across cultures. Therefore, the present study aimed to clarify age-related differences of school-aged children’s ToM and its sociocultural correlates and underlying cognitive mechanisms in two cultural samples of 5.5 - to 12-year-old children from China (n = 126) and Australia (n = 83). The overall project has been broken into three studies. Study 1 assessed Chinese and Australian children’s ToM using four measures, which reflect three categories: cognitive ToM (Interpretive ToM [IToM]), affective ToM (Faux Pas and Reading Mind in the Eyes [RMIE]), and mixed task (Strange Stories). The results revealed that younger children aged from 5.5 to 7 years in both cultural groups performed worse than older children aged from 8 to 11 years on all measures. Australian children outperformed Chinese children on both RMIE and Strange Stories. In contrast, no cultural differences were found in children’s scores on IToM and Faux Pas. However, an analysis of types of errors on the two tasks reflected cultural influences. Chinese children made more reality/contamination errors in the IToM task than Australian children, whereas 10- to 12-year-old Australian children made more identical responses than their Chinese age cohorts. Moreover, Australian children made fewer errors in Faux Pas identification than Chinese children. Study 2 examined how parenting practices and sibling composition influenced ToM across cultures. The results revealed that corporal punishment negative, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Applied Psychology, Griffith Health, Full Text

In large-scale wireless networks, effective transceiver design based on imperfect knowledge of the wireless channel is crucial. In this work, we introduce methods for transceiver design and resource allocation that make use of asymptotic results to reduce computational complexity. We focus on linear transceivers, which are close to optimal for the system dimensions we have in mind. Since the introduced methods rely heavily on second order statistics of the wireless channel, we further discuss efficient methods for the acquisition of these statistics in practical systems., Die Entwicklung von effektiven Algorithmen für Sende- und Empfangsverarbeitung auf Grundlage von imperfekter Kanalinformation spiel eine Schlüsselrolle in hochdimensionalen drahtlosen Kommunikationssystemen. In dieser Arbeit stellen wir entsprechende Algorithmen vor, welche asymptotische Ergebnisse ausnutzen um die Komplexität zu verringern. Wir setzen den Fokus auf lineare Sende- und Emfpangsfilter, da diese für die hochdimensionalen Systeme die wir betrachten nahezu optimal sind. Da wir für unser Filterdesign auf die Verfügbarkeit von statistischer Information über den Kanal angewiesen sind, diskutieren wir des weiteren effiziente Methoden um diese statistische Information zu schätzen.

Deep Neural Networks (DNNs) are known to be strong predictors, but their prediction strategies can rarely be understood. With recent advances in Explainable Artificial Intelligence (XAI), approaches are available to explore the reasoning behind those complex models' predictions. Among post-hoc attribution methods, Layer-wise Relevance Propagation (LRP) shows high performance. For deeper quantitative analysis, manual approaches exist, but without the right tools they are unnecessarily labor intensive. In this software paper, we introduce three software packages targeted at scientists to explore model reasoning using attribution approaches and beyond: (1) Zennit - a highly customizable and intuitive attribution framework implementing LRP and related approaches in PyTorch, (2) CoRelAy - a framework to easily and quickly construct quantitative analysis pipelines for dataset-wide analyses of explanations, and (3) ViRelAy - a web-application to interactively explore data, attributions, and analysis results. With this, we provide a standardized implementation solution for XAI, to contribute towards more reproducibility in our field., Comment: 20 pages, 6 figures, 2 listings, 1 table

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Applied Psychology, Griffith Health, The modern world offers increased connectivity and multiculturism. In this connected world, individuals are frequently granted the opportunity to interact with others from distinct social, ethnic, and religious groups in everyday life. A large body of research suggests that intergroup contact across distinct groups will reduce prejudices and intergroup biases. However, despite the increased opportunity for prejudice- reducing intergroup contact, prejudices remain high in multicultural societies. In exploring this apparent contradiction, recent research suggests that, despite increased opportunities for intergroup contact, many people selectively interact with members from their ingroup. Thus, intergroup avoidance can impede progress in social cohesion and intergroup harmony by preventing beneficial encounters taking place. Previous research has found intergroup anxiety, or the expectation that outgroup members pose a threat to an individual and their ingroup, is strongly and reliably linked to avoidant behaviours. In turn, the most efficacious way to reduce anxiety is through intergroup contact. The problem of intergroup avoidance is therefore circular, where the best way to reduce avoidance is to promote approach behaviours by those who are avoidant. Reducing intergroup anxiety using techniques more likely to be engaged with could interrupt this circular problem, thereby reducing intergroup avoidance. This thesis comprises two theoretical reviews, one systematic review (k = 72), four survey-based studies (N = 2,095) and one experiment (N = 82), that collectively seek to further our understanding of, and provide viable remedies for intergroup anxiety and avoidance to enhance intergroup relations. More specifically, the thesis aims to (1) integrate associative learning and social psychological theories to explain the formation of intergroup anxiety, (2) understand how intergroup anxiety can contribute to a deterioration in intergroup relations, through intergroup avo

Full Text, Thesis (Professional Doctorate), Doctor of Philosophy in Clinical Psychology (PhD ClinPsych), School of Applied Psychology, Griffith Health, Disgust has been implicated in the aetiology and maintenance of various forms of psychopathology, including contamination related obsessive-compulsive disorder (C-OCD). Disgust is difficult to diminish using standard exposure-based interventions. It can be acquired through evaluative conditioning (EC) where a disgust valence towards a neutral stimulus (CS) can be learnt through its contingent presentation with a disgusting stimulus (US). Drawing on the expectancy learning model, exposure-based interventions successfully reduce fear towards the CS by violating danger expectancy, but the affective valence of the CS is less impacted by similar methods. Strategies utilising evaluative conditioning directly target valence of the CS-US relation by repeated presentations of the stimulus of positive valence along with CS (i.e., in counterconditioning) or with US (i.e., in US revaluation). This thesis compared the effects of exposure-based and EC-based strategies in reducing disgust across healthy adults and C-OCD analogue samples. Study 1 compared the effects of exposure in reducing disgust and danger expectancy with EC-based interventions in a sample of students using a picture pairing paradigm. It was found that disgust was acquired through EC, as indicated by both self-report and EMG data. It was further found that at the subjective level, disgust was effectively diminished by all interventions with an exception of a control group. EC-strategies showed larger effect sizes in reducing disgust, while exposure-based interventions showed greater effect sizes in reducing US expectancy. Only counterconditioning reduced EMG levator activity associated with disgust at extinction phase. However, participants in this group showed increased scores on contamination and checking OCD subscales. After controlling for these variables, EMG levator activity remained increased at extinction but unaffected by the intervention. The findings suggested possible differential disgust conditionin

The purpose of the Impacts on Bear Lake Storage under Alternative High-Runoff Management Operations study is to identify how much additional water could have been stored in Bear Lake between 1980 and 2018 if changes were made to the criteria by which PacifiCorp operates Bear Lake to reduce the risk of flooding on Bear Lake and on the Bear River below Bear Lake. Besides storage in Bear Lake, effects of altering operating policy are evaluated at other locations in the Lower Bear River system, including the Causeway, Gentile Valley, downstream of Oneida reservoir, downstream of Cutler reservoir, and inflow to Great Salt Lake. This executive summary provides a brief description of the study background, scenarios analyzed, key findings, model limitations, and recommendations for further study and model development.

The purpose of the Impacts on Bear Lake Storage under Alternative High-Runoff Management Operations study is to identify how much additional water could have been stored in Bear Lake between 1980 and 2018 if changes were made to the criteria by which PacifiCorp operates Bear Lake to reduce the risk of flooding on Bear Lake and on the Bear River below Bear Lake. Besides storage in Bear Lake, effects of altering operating policy are evaluated at other locations in the Lower Bear River system, including the Causeway, Gentile Valley, downstream of Oneida reservoir, downstream of Cutler reservoir, and inflow to Great Salt Lake. This executive summary provides a brief description of the study background, scenarios analyzed, key findings, model limitations, and recommendations for further study and model development.

Digital contact tracing approaches based on Bluetooth low energy (BLE) have the potential to efficiently contain and delay outbreaks of infectious diseases such as the ongoing SARS-CoV-2 pandemic. In this work we propose a novel machine learning based approach to reliably detect subjects that have spent enough time in close proximity to be at risk of being infected. Our study is an important proof of concept that will aid the battery of epidemiological policies aiming to slow down the rapid spread of COVID-19.

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Applied Psychology, Griffith Health, While the benefits of participating in regular leisure-time exercise are well known, many individuals fail to exercise regularly. Contemporary research into exercise motivation has identified automatic cognitive processes, such approach-avoidance biases, as at least partly involved in motivating exercise behaviour. Approach-avoidance biases refer to automatic and implicit dispositions to approach or avoid certain cues in the environment. These biases can be examined by measuring immediate action tendencies to approach or avoid, or by examining implicit approach-avoidance associations stored in long-term memory. Specifically, approach-avoidance associations reflect pre-existing links in memory between a target concept (e.g., exercise) and behavioural tendencies to approach or avoid. To date, existing research has not yet demonstrated a link between implicit approach-avoidance associations with exercise cues and exercise behaviour. Identifying whether participation in leisure-time exercise is associated with the strength of approach-avoidance associations may provide insight into why some individuals are automatically motivated to pursue exercise as a form of leisure and others are not. Therefore, the overarching aim of this thesis was to contribute to current understanding of exercise motivation by investigating the relationship between approach-avoidance memory associations and leisure-time exercise engagement. Study 1 used a Recoding-Free Implicit Association Test (IAT-RF) to examine the relationship between self-reported leisure-time exercise engagement and implicit approach-avoidance memory associations with exercise in a sample of healthy adults (N = 104). Results revealed stronger exercise-approach versus exercise-avoidance memory associations, indicative of an implicit approach bias for exercise cues. In addition, approach bias scores positively correlated with self-reported time spent on leisure-time exercise, and discriminated between individuals who did or

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Applied Psychology, Griffith Health, Sex differences in cognitive ability have been documented in psychological research for over a century, and the research area has seen considerable changes in theoretical perspectives and methodology. While males and females do not differ in general intelligence, an extensive body of literature documents sex differences in more specific cognitive tasks (for reviews see Halpern, 2000; Kimura, 2000; Maccoby & Jacklin, 1974). Males on average perform at a higher level on tasks that rely on visualspatial ability, and this has been linked to later gender gaps in quantitative abilities such as mathematics and science and to the underrepresentation of women in science, technology, engineering and mathematics (STEM)-related fields. Females as a group do better at tasks involving verbal and language abilities which have been linked to wide gender gaps in reading and writing, as well as the underrepresentation of men in postsecondary education. Some researchers have argued that sex differences in cognitive ability are declining in response to social changes in the roles and status of women, but methodological limitations and use of convenience samples have limited previous enquiries seeking to test that hypothesis. The aim of this course of research was twofold. Firstly, using the statistical technique of meta-analysis to examine the evidence for sex differences in visual-spatial, verbal and quantitative abilities, and - given the passage of time - whether they were declining in response to changes in the roles of men and women in society. This was addressed through a series of studies that examined: i) nationally representative samples of student testing data from the National Assessment of Educational Progress (NAEP) in the United States, ii) cross-cultural samples of student testing data from the Programme for International Student Assessment (PISA). Secondly, to determine the contribution of sex-typed personality traits and behaviours (collectively referred to as sex-role

Contemporary learning models for computer vision are typically trained on very large (benchmark) datasets with millions of samples. These may, however, contain biases, artifacts, or errors that have gone unnoticed and are exploitable by the model. In the worst case, the trained model does not learn a valid and generalizable strategy to solve the problem it was trained for, and becomes a 'Clever-Hans' (CH) predictor that bases its decisions on spurious correlations in the training data, potentially yielding an unrepresentative or unfair, and possibly even hazardous predictor. In this paper, we contribute by providing a comprehensive analysis framework based on a scalable statistical analysis of attributions from explanation methods for large data corpora. Based on a recent technique - Spectral Relevance Analysis - we propose the following technical contributions and resulting findings: (a) a scalable quantification of artifactual and poisoned classes where the machine learning models under study exhibit CH behavior, (b) several approaches denoted as Class Artifact Compensation (ClArC), which are able to effectively and significantly reduce a model's CH behavior. I.e., we are able to un-Hans models trained on (poisoned) datasets, such as the popular ImageNet data corpus. We demonstrate that ClArC, defined in a simple theoretical framework, may be implemented as part of a Neural Network's training or fine-tuning process, or in a post-hoc manner by injecting additional layers, preventing any further propagation of undesired CH features, into the network architecture. Using our proposed methods, we provide qualitative and quantitative analyses of the biases and artifacts in various datasets. We demonstrate that these insights can give rise to improved, more representative and fairer models operating on implicitly cleaned data corpora., Comment: 47 pages, 21 figures

The field of video compression has developed some of the most sophisticated and efficient compression algorithms known in the literature, enabling very high compressibility for little loss of information. Whilst some of these techniques are domain specific, many of their underlying principles are universal in that they can be adapted and applied for compressing different types of data. In this work we present DeepCABAC, a compression algorithm for deep neural networks that is based on one of the state-of-the-art video coding techniques. Concretely, it applies a Context-based Adaptive Binary Arithmetic Coder (CABAC) to the network's parameters, which was originally designed for the H.264/AVC video coding standard and became the state-of-the-art for lossless compression. Moreover, DeepCABAC employs a novel quantization scheme that minimizes the rate-distortion function while simultaneously taking the impact of quantization onto the accuracy of the network into account. Experimental results show that DeepCABAC consistently attains higher compression rates than previously proposed coding techniques for neural network compression. For instance, it is able to compress the VGG16 ImageNet model by x63.6 with no loss of accuracy, thus being able to represent the entire network with merely 8.7MB. The source code for encoding and decoding can be found at https://github.com/fraunhoferhhi/DeepCABAC.

We present DeepCABAC, a novel context-adaptive binary arithmetic coder for compressing deep neural networks. It quantizes each weight parameter by minimizing a weighted rate-distortion function, which implicitly takes the impact of quantization on to the accuracy of the network into account. Subsequently, it compresses the quantized values into a bitstream representation with minimal redundancies. We show that DeepCABAC is able to reach very high compression ratios across a wide set of different network architectures and datasets. For instance, we are able to compress by x63.6 the VGG16 ImageNet model with no loss of accuracy, thus being able to represent the entire network with merely 8.7MB., Comment: ICML 2019, Joint Workshop on On-Device Machine Learning and Compact Deep Neural Network Representations (ODML-CDNNR)

Full Text, Thesis (Professional Doctorate), Doctor of Philosophy in Clinical Psychology (PhD ClinPsych), School of Applied Psychology, Griffith Health, Cognitive models of social anxiety disorder (SAD) emphasize repetitive negative thinking as a prominent maintaining factor that occurs both before (anticipatory processing) and after (post-event processing) social-evaluative events. While evidence suggests that repetitive negative thinking is a maladaptive process that has damaging cognitive, behavioural, and affective consequences for those with SAD, the cognitive mechanisms thought to underlie ineffective control of attention on these modes of thought was unclear. Cognitive interference theories, such as attentional control theory, propose that high levels of anxiety negatively impact the processing efficiency on tasks that require high cognitive resources. Accordingly, this thesis aimed to elucidate the role of executive functions on repetitive negative thinking in social anxiety by employing a combination of self-report and experimental methodologies with social anxiety analogue samples. Studies 1A and 1B explored the correlational relationships between social anxiety, anticipatory processing and post-event processing, and attentional control, as an initial step prior to conducting experimental studies. Poorer total attentional control was associated with social anxiety in both studies and diminished attentional shifting was associated with increased levels of PEP, but not AP. Mediation modelling suggested that trait anxiety mediated the relationship between attentional shifting and PEP, and positive affect mediated the relationship between total attentional control and PEP. Drawing from attentional control theory, studies 2 and 3, manipulated both anticipatory processing (study 2) and post-event processing (study 3), to examine the underlying executive functions responsible for ineffective control of attention in anxiety (i.e., inhibition and shifting) using the mixed emotional saccade task. Both studies 2 and 3 did not demonstrate an anxiety-related effect on attentional control and were comparable in showing

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Applied Psychology, Griffith Health, Empathy is an essential social communication skill for sharing and understanding others’ emotional states and experiences. It is deemed to be related to both cultural and personal factors. A few researchers have investigated Western–Asian cross-cultural differences in empathy using self-report questionnaires or behavioural tasks, and some of them found that cultural differences in empathy were evident. Nevertheless, the number of these studies is small and the results reported are inconsistent. Furthermore, reasons underlying the cultural differences have hardly been investigated. The current thesis was conducted to address the limitations in the literature (Chapter II), to bridge some theoretical gaps in the research area (Chapter II), and to provide a better understanding of Western–Asian cross-cultural differences in both self-report and behavioural responses of empathy using Australian and Mainland Chinese participants (Chapters V and VI). Three studies were conducted. In the first study (Chapter IV), the Empathy Quotient (EQ) was validated in a sample of participants (n = 588) from Mainland China. Results confirmed the validity of the EQ for measuring self-report empathy in the Mainland Chinese population. Moreover, with a comparison with previous publications based on Western populations, the mean score of the EQ was found to be lower in the Mainland Chinese participants; the sex differences in the EQ score was found to be smaller in Mainland Chinese participants, relative to Westerners. Furthermore, the best-fit model of the EQ for Mainland Chinese participants was a one-factor model, suggesting that psychometric characteristics of items for measuring emotional empathy and cognitive empathy were undifferentiated, which was different from the dissociable model reported for Western populations. The possible cultural differences in the mean value of the EQ scale and in the effect size of sex differences in the EQ score revealed in the first study implied a cultu

The present study tested whether practice using a performance goal enhances accuracy when later tested under the demands of a competition outcome goal for a basketball free throw task. Thirty male elite-level basketball players with a mean age of 24.67 years practiced free throws either with or without a performance goal. The performance goal was constructed with the challenge of bettering an individualised accuracy level. Players were later tested under an outcome goal that emphasised performance in an external competitive situation. Performance was best in players that practiced using performance goals and greater gains were seen in players that showed an initial lower level of performance. Practice using performance goals may improve free throw accuracy in a game situation, particularly for players that are at a relatively lower skill level.

The present study tested whether practice using a performance goal enhances accuracy when later tested under the demands of a competition outcome goal for a basketball free throw task. Thirty male elite-level basketball players with a mean age of 24.67 years practiced free throws either with or without a performance goal. The performance goal was constructed with the challenge of bettering an individualised accuracy level. Players were later tested under an outcome goal that emphasised performance in an external competitive situation. Performance was best in players that practiced using performance goals and greater gains were seen in players that showed an initial lower level of performance. Practice using performance goals may improve free throw accuracy in a game situation, particularly for players that are at a relatively lower skill level.

Exposure therapy, which is based on extinction in Pavlovian conditioning, is effective in the treatment of anxiety disorders. However, a return of fear (relapse) can occur after treatment. Reinstatement and spontaneous recovery provide two explanations for return of fear. These mechanisms can be difficult to investigate in real-world clinical contexts for practical or ethical reasons. The present research used a laboratory-based fear learning task to examine reinstatement and spontaneous recovery using different measures of fear learning. A differential fear conditioning procedure was used. In acquisition, one conditional stimulus (CS+) was paired with a shock unconditional stimulus (US) and a CS- was presented alone. Both CS+ and CS- were presented alone during extinction. Presentations of the US (reinstatement) or a time delay (spontaneous recovery) were given prior to test trials. During test, shock expectancy was greater for CS+ than CS- for reinstatement and was non-differential, although higher than at the end of extinction, for spontaneous recovery. The CS+ was rated as more unpleasant, more arousing, and more dominating than the CS- for both procedures. Skin conductance responses did not differ between the CS+ and CS-. The results show dissociations between the measures that may reflect differential sensitivity to the learning processes that underlie the return of fear. The application of the results for understanding the possible mechanisms for relapse following exposure therapy for anxiety disorders is discussed.

Individuals can focus their attention in different ways during exercise and different foci may influence psychological states experienced. The present study examined the distribution of attentional focus strategies in exercisers and their relationships with gender, perceived exertion, enjoyment, and satisfaction. Regular exercisers (176 females, 144 males) completed a measure of attentional focus and rated their perceived exertion, enjoyment, and satisfaction during an exercise session. All participants used more than one type of attentional focus during an exercise session. Males spent more time than females attending to task-relevant thoughts and external cues whereas females spent more time attending to task-irrelevant thoughts and external distractions. In females and males, time spent engaging in task-irrelevant thoughts was negatively correlated with perceived exertion and satisfaction with exercise. For females only, time using external distractions was negatively correlated with satisfaction and positively correlated with enjoyment. For males only, time attending to task-relevant external cues was positively correlated with perceived exertion and enjoyment. The observed gender differences in attentional focus preferences and the relationships with psychological states have implications for advice given to exercisers and approaches that aim to promote adherence to exercise programs.

Individuals can focus their attention in different ways during exercise and different foci may influence psychological states experienced. The present study examined the distribution of attentional focus strategies in exercisers and their relationships with gender, perceived exertion, enjoyment, and satisfaction. Regular exercisers (176 females, 144 males) completed a measure of attentional focus and rated their perceived exertion, enjoyment, and satisfaction during an exercise session. All participants used more than one type of attentional focus during an exercise session. Males spent more time than females attending to task-relevant thoughts and external cues whereas females spent more time attending to task-irrelevant thoughts and external distractions. In females and males, time spent engaging in task-irrelevant thoughts was negatively correlated with perceived exertion and satisfaction with exercise. For females only, time using external distractions was negatively correlated with satisfaction and positively correlated with enjoyment. For males only, time attending to task-relevant external cues was positively correlated with perceived exertion and enjoyment. The observed gender differences in attentional focus preferences and the relationships with psychological states have implications for advice given to exercisers and approaches that aim to promote adherence to exercise programs.

We present a novel approach for low-complexity equalizer design well-suited for cellular massive MIMO systems. Our design allows to exploit the channel structure in terms of covariance matrices to improve the performance in the face of pilot-contamination, while basically keeping the complexity of a matched filter. This is achieved by restricting the equalizer to functions which are bilinear in the received data signals and the observations from a training phase. The proposed design generalizes several previous approaches to equalizer design for massive MIMO. We show by asymptotic analysis that with the proposed design the achievable rate grows without bound for growing numbers of antennas even in the presence of pilot-contamination. We demonstrate with numerical results that the proposed design is competitive with more complex approaches in a practical cellular setup.

We present a method for estimating conditionally Gaussian random vectors with random covariance matrices, which uses techniques from the field of machine learning. Such models are typical in communication systems, where the covariance matrix of the channel vector depends on random parameters, e.g., angles of propagation paths. If the covariance matrices exhibit certain Toeplitz and shift-invariance structures, the complexity of the MMSE channel estimator can be reduced to O(M log M) floating point operations, where M is the channel dimension. While in the absence of structure the complexity is much higher, we obtain a similarly efficient (but suboptimal) estimator by using the MMSE estimator of the structured model as a blueprint for the architecture of a neural network. This network learns the MMSE estimator for the unstructured model, but only within the given class of estimators that contains the MMSE estimator for the structured model. Numerical simulations with typical spatial channel models demonstrate the generalization properties of the chosen class of estimators to realistic channel models., Comment: To appear in IEEE Transactions on Signal Processing

Interference during the uplink training phase significantly deteriorates the performance of a massive MIMO system. The impact of the interference can be reduced by exploiting second order statistics of the channel vectors, e.g., to obtain minimum mean squared error estimates of the channel. In practice, the channel covariance matrices have to be estimated. The estimation of the covariance matrices is also impeded by the interference during the training phase. However, the coherence interval of the covariance matrices is larger than that of the channel vectors. This allows us to derive methods for accurate covariance matrix estimation by appropriate assignment of pilot sequences to users in consecutive channel coherence intervals., Comment: submitted to IEEE Signal Processing Letters

The main aim of the study was to evaluate maternal and newborn urinary iodine concentrations according to the usage of iodine supplementation during pregnancy., The main aim of the study was to evaluate maternal and newborn urinary iodine concentrations according to the usage of iodine supplementation during pregnancy.

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy in Clinical Psychology (PhD ClinPsych), School of Applied Psychology, Griffith Health, Social cognition refers to the collection of cognitive processes that enable fluid social engagement. One of the core processes subsumed under social cognition is theory of mind (ToM), which refers to the capacity to infer the thoughts, feelings, and intentions of others. A substantial body of research has demonstrated that ToM is impaired in a variety of neuropsychiatric disorders including schizophrenia. There is good empirical evidence that many psychotic symptoms may best be understood in light of a disturbed capacity to monitor one’s own and others’ intentions. Nonetheless, exactly how impaired ToM contributes to the clinical presentation of schizophrenia remains unclear, as is the magnitude of deterioration in ToM that results from illness chronicity. Furthermore, the extent to which ToM is associated with neurocognition and other domains of social cognition (e.g., empathy) is still under debate. Preliminary research has also started to examine the different types and functional consequences of ToM impairment in schizophrenia. Examination of these potential research areas has likely been limited by the lack of psychometrically sound and sensitive measures of ToM that are suitable for use with healthy and clinical adult populations.

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Applied Psychology, Griffith Health, A relationship exists between mood and cardiac control systems. This relationship has been established through correlations between medical pathology, such as Coronary Heart Disease (CHD), and psychopathological changes in mood, such as Major Depressive Disorder (MDD). Euthymic mood and normal cardiac regulation, as well as the diseases MDD and CHD, are linked by numerous interrelated physiological pathways. These pathways form part of a physiological regulatory network in the body called the psycho-immune-neuroendocrine (PINE) network. PINE network homeostasis can be disrupted by stress, resulting in pathological changes, which give rise to MDD, CHD, or both diseases. Heart Rate Variability (HRV) is a physiological measure of cardiac function which reflects autonomic cardiac control, an important component of the PINE network. This body of work set out to test three hypotheses, the first being that psychometric measures of mood and HRV measures are related across a broad range of mood, from non-pathological to pathological depressed mood. The second hypothesis is that psychometric measures of mood and HRV measures remain related over time, even when an individual’s mood changes across time. The third hypothesis is that in people with CHD, HRV measures would still be related to mood, but HRV may be globally reduced compared to people without CHD. A longitudinal cohort study, the Heart and Mind Study, recruited 88 participants and followed them up over six months.

We introduce novel blind and semi-blind channel estimation methods for cellular time-division duplexing systems with a large number of antennas at each base station. The methods are based on the maximum a-posteriori principle given a prior for the distribution of the channel vectors and the received signals from the uplink training and data phases. Contrary to the state-of-the-art massive MIMO channel estimators which either perform linear estimation based on the pilot symbols or rely on a blind principle, the proposed semi-blind method efficiently suppresses most of the interference caused by pilot-contamination. The simulative analysis illustrates that the semi-blind estimator outperforms state- of-the-art linear and non-linear approaches to the massive MIMO channel estimation problem.

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy in Clinical Psychology (PhD ClinPsych), School of Applied Psychology, Griffith Health, The anxiety disorders are a collection of disorders characterized by psychological fear and anxiety and somatic manifestations of fear and anxiety (4th ed., text rev.; DSM-IV-TR; Diagnostic and Statistical Manual of Mental Disorders; American Psychiatric Association, 2000). Anxiety disorders can significantly impair a sufferers overall wellbeing and result in significant individual and societal economic burden. Exposure therapy has been proposed to be a core component of anxiety disorder treatment. However, up to 50% of those who are successfully treated experience return of fear after therapy, which may act as a precursor of complete clinical relapse. Spontaneous recovery, reacquisition, reinstatement, and renewal may be four contextually driven mechanisms that can predict return of fear. Renewal, the most widely studied of these mechanisms, was the predominant focus of the current thesis. Renewal of fear occurs when post-therapy encounters with feared stimuli take place in contexts that are different to the treatment context.

Systematic review methodology can be distinguished from narrative reviews of the literature through its emphasis on transparent, structured and comprehensive approaches to searching the literature and its requirement for formal synthesis of research findi

Full Text, Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Education and Professional Studies, Arts, Education and Law, The execution of skilled performance under the pressure of competition is the ultimate challenge of sports performers. The attentional focus adopted when performing under pressure can be a key factor in success. Additionally, the skill level of the athlete may be a mediating factor in the optimal focus of attention. Research on skill acquisition suggests that elite athletes are likely to be operating at an automatic level of execution, whereas novice athletes are more likely to require cognitive resources during performance. Previous research in endurance sports has indicated that athletes will benefit during competition from an association focus, although this may be a combination of internal and external cues. Cycling is one endurance sport that has received considerable attention from research disciplines such as physiology and biomechanics, but very little in the way of psychology. Just like other sports, the focus of attention that cyclists adopt when riding may have a significant impact on their performance. This thesis therefore aimed to investigate the optimal focus of attention of cyclists of all abilities, and particularly for elite athletes under the pressure of competition.

The renewal of Pavlovian-conditioned responses may provide a model for the relapse of fear following extinction-based treatments for anxiety disorders. Renewal can be observed if conditional stimulus (CS) and unconditional stimulus (US) pairings are given in one context, extinction trials of CS presentations in a second context, prior to test trials of CS presentations in the original acquisition context (ABA renewal). We examined ABA renewal in humans by using a fear-conditioning procedure with an unpleasant shock US. A renewal of rated shock expectancy was demonstrated with this procedure. Conducting extinction treatment in multiple contexts was expected to attenuate the renewal effect. However, the renewal of shock expectancy persisted when extinction treatment was given across three or five different contexts. With the current renewal design, learning task, and measure of conditioned behaviour, extinction treatment does not appear to readily generalise to the test context. The use of multiple extinction treatments in a clinical setting may not necessarily reduce the likelihood of relapse via a renewal effect.

Using a 14-mm thick volume Bragg grating, spectral bandwidth of a diode laser array is narrowed to 7 GHz (FWHM). The total output power is 13.5 W, of which 86% is in the 7-GHz band. With such a narrow bandwidth, it is possible to temperature tune laser frequency across O(2) X(3) Pi -> b(1) Sigma(+) absorption line at 763.8 nm, efficiency generating O(2)((1)Delta) molecules., Pub. in Optical Society of America, 2006.

Prepulse inhibition of the blink reflex is widely applied to investigate information processing deficits in schizophrenia and other psychiatric patient groups. The present experiment investigated the hypothesis that prepulse inhibition reflects a transient process that protects preattentive processing of the prepulse. Participants were presented with pairs of blink-eliciting noises, some preceded by a prepulse at a variable stimulus onset asynchrony (SOA), and were asked to rate the intensity of the second noise relative to the first. Inhibition of blink amplitude was greater for a 110-dB(A) noise than for a 95-dB(A) noise with a 120-ms SOA, whereas there was no difference with a 30-ms SOA. The perceived intensity was also lower for the 110-dB(A) noise than for the 95-dB(A) noise with the 120-ms SOA, but not with the 30-ms SOA. The parallel results support a relationship between prepulse inhibition of response amplitude and perceived intensity. However, the prepulse did not reduce intensity ratings relative to control trials in some conditions, suggesting that prepulse inhibition is not always associated with an attenuation of the perceived impact of the blink-eliciting stimulus.

The blink reflex is modulated if a weak lead stimulus precedes the blink-eliciting stimulus. In two experiments, we examined the effects of the sensory modality of the lead and blink-eliciting stimuli on blink modulation. Acoustic, visual, or tactile lead stimuli were followed by an acoustic (Experiment 1) or an electrotactile (Experiment 2) blink-eliciting stimulus at lead intervals of -30, 0, 30, 60, 120, 240, 360, and 4,500 msec. The inhibition of blink magnitude at the short (60- to 360-msec) lead intervals and the facilitation of blink magnitude at the long (4,500-msec) lead interval observed for each lead stimulus modality was relatively unaffected by the blink-eliciting stimulus modality. The facilitation of blink magnitude at the very short ( -30- to 30-msec) lead intervals was dependent on the combination of the lead and the blink-eliciting stimulus modalities. Modality specific and nonspecific processes operate at different levels of perceptual processing.

The effects of the sensory modality of the lead stimulus and of task difficulty on attentional modulation of the electrical and acoustic blink reflex were examined. Participants performed a discrimination and counting task with either two acoustic, two visual, or two tactile lead stimuli. In Experiment 1, facilitation of the electrically elicited blink was greater during task-relevant than during task-irrelevant lead stimuli. Increasing task difficulty enhanced magnitude facilitation for acoustic lead stimuli. In Experiment 2, acoustic blink facilitation was greater during task-relevant lead stimuli, but was unaffected by task difficulty. Experiment 3 showed that a further increase in task difficulty did not affect acoustic blink facilitation during visual lead stimuli. The observation that blink reflexes are facilitated by attention in the present task domain is consistent across a range of stimulus modality and task difficulty conditions.

Four experiments investigated the attentional modulation of acoustic blinks during continuous spatial tracking tasks. Experiment 1 found blink magnitude inhibition in a visual tracking task. Experiment 2 replicated this finding and also found blink latency slowing. Experiment 3 varied the difficulty of the task and found larger blink inhibition in the easy condition. Blink latency slowing did not differ and was significant at both difficulty levels. Experiment 4 employed less difficult visual and acoustic tracking tasks at two levels of task load. Blink magnitude inhibition during the visual and facilitation during the acoustic task was significant during high load in both modality groups. Blink latency was slowed in all visual task conditions and shortened in the difficult acoustic task. These results indicate that attentional blink modulation in a continuous spatial tracking task is modality specific.

The effects of attention to a lead stimulus and of its sensory properties on modulation of the acoustic blink reflex were investigated. Participants performed a reaction time task cued by an acoustic or a visual lead stimulus. In Experiment 1, half the participants were presented with sustained lead stimuli. For the remainder, the lead stimulus was discrete and consisted of two brief presentations that marked the onset and offset of a stimulus-free interval. In Experiment 2, sustained lead stimuli were presented at a low or high intensity. The attentional demands of the task enhanced blink latency and magnitude modulation during acoustic and visual lead stimuli, with blink modulation being largest at a late point during the lead stimulus. Independent of the attentional effects, blink latency and magnitude modulation were larger during sustained than during discrete acoustic lead stimuli, whereas there was no difference for visual lead stimuli. Increases in the intensity of the lead stimulus enhanced blink modulation regardless of lead stimulus modality. Attention to a lead stimulus and the properties of the lead stimulus appear to have independent effects on blink reflex modulation.

Previous studies found larger attentional modulation of acoustic blinks during task-relevant than during task-irrelevant acoustic or visual, but not tactile, lead stimuli. Moreover, blink modulation was larger overall during acoustic lead stimuli. The present experiment investigated whether these results reflect modality specificity of attentional blink modulation or effects of continuous stimulation. Participants performed a discrimination and counting task with acoustic, visual, or tactile lead stimuli. Stimuli were presented sustained or consisted of two short discrete stimuli. The sustained condition replicated previous results. In the discrete condition, blinks were larger during task-relevant than during task-irrelevant stimuli in all groups regardless of lead stimulus modality. Thus, previous results that seemed consistent with modality-specific accounts of attentional blink modulation reflect effects of continuous stimulus input.

Two experiments investigated the effects of the sensory modality of the lead and of the blink-eliciting stimulus during lead stimulus modality change on blink modulation at lead intervals of 2500 and 3500 ms. Participants were presented with acoustic, visual, or tactile change stimuli after habituation training with lead stimuli from the same or a different sensory modality. In Experiment 1, latency and magnitude of the acoustic blink were facilitated during a change to acoustic or visual lead stimuli, but not during a change to tactile lead stimuli. After habituation to acoustic lead stimuli, blink magnitude was smaller during tactile change stimuli than during habituation stimuli. The latter finding was replicated in Experiment 2 in which blink was elicited by electrical stimulation of the trigeminal nerve. The consistency of the findings across different combinations of lead stimulus and blink-eliciting stimulus modalities does not support a modality-specific account of attentional blink modulation. Rather, blink modulation during generalized orienting reflects modality non-specific processes, although modulation may not always be found during tactile lead stimuli.

Measures of eye activity, such as blink rate and scanning patterns, have been used extensively as psychophysiological indices of mental workload. In a review of measures derived from spontaneous eye activity it is shown that different measures are differentially sensitive to specific aspects of mental workload. A less well-known measure of non-spontaneous eye activity, the blink reflex, is also reviewed. Experiments using discrete punctuate stimuli and continuous tasks analogous to real-world systems show that blink reflexes are modulated by attention and that this modulation reflects modality-specific attentional engagement. Future research should examine the utility of the blink reflex according to the desirable properties of sensitivity, diagnosticity, validity, reliability, ease of use, unobtrusiveness, and operator acceptance.

The effect that the difficulty of the discrimination between task-relevant and task-irrelevant stimuli has on the relationship between skin conductance orienting and secondary task reaction time (RT) was examined. Participants (N = 72) counted the number of longer-than-usual presentations of one shape (task-relevant) and ignored presentations of another shape (task-irrelevant). The difficulty of discriminating between the two shapes varied across three groups (low, medium, and high difficulty). Simultaneous with the primary counting task, participants performed a secondary RT task to acoustic probes presented 50, 150, and 2000 ms following shape onset. Skin conductance orienting was larger, and secondary RT at the 2000 ms probe position was slower during task-relevant shapes than during task-irrelevant shapes in the low-difficulty group. This difference declined as the discrimination difficulty was increased, such that there was no difference in the high-difficulty group. Secondary RT was slower during task-irrelevant shapes than during task-relevant shapes only in the medium-difficulty group - and only at the 150 ms probe position in the first half of the experiment. The close relationship between autonomic orienting and secondary RT at the 2000 ms probe position suggests that orienting reflects the resource allocation that results from the number of matching features between a stimulus input and a mental representation primed as significant.

An optically pumped, For laser was demonstrated operating at 4.0 micrometers. This is the first demonstration of an HBr laser by direct optical pumping of the 0 right arrow 3 vibrational overtone band at 1.34 micrometers. An injection locked, temperature tuned Nd: YAG laser was used as the optical pump source. Lasing was achieved in a cell of HRr maintained at 100 Torr. potential for scaling to higher pulse energies has been demonstrated. The HBr laser has potential use as a source for infrared countermeasures., Prepared in collaboration with the University of New Mexico, Dept. of Physics and Astronomy, Albuquerque, NM.

Evaluative learning theory states that affective learning, the acquisition of likes and dislikes, is qualitatively different from relational learning, the learning of predictive relationships among stimuli. Three experiments tested the prediction derived from evaluative learning theory that relational learning, but not affective learning, is affected by stimulus competition by comparing performance during two conditional stimuli, one trained in a superconditioning procedure and the other in a blocking procedure. Ratings of unconditional stimulus expectancy and electrodermal responses indicated stimulus competition in relational learning. Evidence for stimulus competition in affective learning was provided by verbal ratings of conditional stimulus pleasantness and by measures of blink startle modulation. Taken together, the present experiments demonstrate stimulus competition in relational and affective learning, a result inconsistent with evaluative learning theory.

Information processing accounts propose that autonomic orienting reflects the amount of resources allocated to process a stimulus. However, secondary task reaction time (RT), a supposed measure of processing resources, has shown a dissociation from autonomic orienting. The present study tested the hypothesis that secondary task RT reflects a serial processing mechanism. Participants (N = 24) were presented with circle and ellipse shapes and asked to count the number of longer-than-usual presentations of one shape (task-relevant) and to ignore presentations of a second shape (task-irrelevant). Concurrent with the counting task, participants performed a secondary RT task to an auditory probe presented at either a high or low intensity and at two different probe positions following shape onset (50 and 300 ms). Electrodermal orienting was larger during task-relevant shapes than during task-irrelevant shapes, but secondary task RT to the high-intensity probe was slower during the latter. In addition, an underadditive interaction between probe stimulus intensity and probe position was found in secondary RT. The findings are consistent with a serial processing model of secondary RT and suggest that the notion of processing stages should be incorporated into current information-processing models of autonomic orienting.