Search

Humans regularly reason from visual information, engaging in simple object search in a scene to abstract mathematical thinking. In recent decades, the field of machine learning has extensively focused on visual tasks with the aim to model human visual reasoning. However, machine learning approaches still do not match human performance on simple visual tasks such as the Synthetic Visual Reasoning Test (SVRT; Fleuret et al. 2011). While this set of tasks is trivial for humans to solve, the current state-of-the-art machine learning algorithms struggle with the SVRT. We argue that the reason for the difference in human reasoning and machines' performance in the SVRT is the ways humans and machines represent the world and visual information specifically. We argue that humans represent situations in terms of relations between constituent objects, and that our representation of these relations is structured and symbolic. By consequence, humans engage in operations that are not available for machine systems that rely on non-structured representations. We hold that operations over structured relational representations is what underlie phenomena such as abstract visual reasoning and cross-domain generalisation. The current work builds on the DORA (Discovery Of Relations by Analogy; Doumas et al., 2008; 2022) model of relation learning. DORA learns structured representations of magnitude relations from simple visual inputs. Here we expand the model to learn more complex categorical relations (e.g., contains or supports) as compressions of simpler relations (e.g., above, in-contact), and develop a new method for identifying relevant relations over which to perform reasoning from simple scenes. We embed the resulting model in a pipeline for human visual reasoning consisting of successful psychological models of object recognition and analogy making. The result is an end-to-end system which is constrained as much as possible by what is known about the processes and mechanisms of the cognitive system-from early vision to learning complex relations and reasoning. The model is tested within the context of the SVRT. The limitations of the model and the directions for future research are discussed.

Relational reasoning is central to many cognitive processes, ranging from "lower" processes like object recognition to "higher" processes such as analogy-making and sequential decision-making. The first chapter of this thesis gives an overview of relational reasoning and the computational demands that it imposes on a system that performs relational reasoning. These demands are characterized in terms of the binding problem in neural networks. There has been a longstanding debate in the literature regarding whether neural network models of cognition are, in principle, capable of relation-base processing. In the second chapter I investigated the relational reasoning capabilities of the Story Gestalt model (St. John, 1992), a classic connectionist model of text comprehension, and a Seq-to-Seq model, a deep neural network of text processing (Bahdanau, Cho, & Bengio, 2015). In both cases I found that the purportedly relational behavior of the models was explainable by the statistics of their training datasets. We propose that both models fail at relational processing because of the binding problem in neural networks. In the third chapter of this thesis, I present an updated version of the DORA architecture (Doumas, Hummel, & Sandhofer, 2008), a symbolic-connectionist model of relation learning and inference that uses temporal synchrony to solve the binding problem. We use this model to perform relational policy transfer between two Atari games. Finally, in the fourth chapter I present a model of relational reinforcement that is able to select relevant relations, from a potentially large pool of applicable relations, to characterize a problem and learn simple rules from the reward signal, helping to bridge the gap between reinforcement learning and relational reasoning.

Evidence suggests music listening and personality can protect against the negative repercussions of everyday stress. Though there is extensive literature on the role of music listening and personality for stress management, genres and traits perceived to be socially aversive have been under-researched in relation to stress, possibly due to their perceived undesirability outside this context. Therefore, their contribution to stress-coping is not yet fully understood. This thesis investigated the role of music genres and personality traits perceived to be socially aversive in the context of stress. Specifically, it (1) assessed the effects of hiphop/rap and the Dark Triad traits (DT; psychopathy, narcissism, Machiavellianism) on stress outcomes, whilst exploring interactions between music, personality, and stress variables, and (2) examined public attitudes towards hip-hop/rap and the DT traits in relation to stress, in nonclinical adult populations. The latter aim was established to assess how receptive the general population would be to these factors' potential stress-coping benefits. Two systematic reviews and a series of laboratory- and survey-based experimental studies revealed that: (a) listening to hip-hop/rap can consistently reduce induced stress irrespective of the experimental environment, stressor type, and listeners' negative attitudes towards its association with stress; (b) narcissism can aid stress management despite the experimental environment and stressor type, though other factors (e.g., music preferences and uses) may influence the manifestation of its stress-coping properties; and (c) establishing hip-hop/rap and narcissism as assets for stress-coping may be received by the general population with less resistance than anticipated, provided that they are publicly depicted as complex and contextually beneficial. These findings suggest that hip-hop/rap and narcissism could be valuable assets for stress-coping, despite their perceived undesirability. By searching 'beneath the surface' of well-established negative connotations, this thesis paves the way for new valuable insights into how music and personality can enhance the well-being and everyday functioning of healthy adults.

Parkinson's is a progressive neurodegenerative condition characterized by a variety of motor and non-motor symptoms that can lead to increasing disability and complex psychosocial challenges. While medical treatments can help people living with Parkinson's to manage symptoms, they do not fully address many challenges associated with the condition, including balance issues, stigma and unpredictability. Art and exercise are increasingly being recognized as therapeutic for people living with Parkinson's, and dance interventions have been demonstrated to improve several motor and non-motor symptoms and quality of life (QOL) in a number of small studies. Outside of laboratory and clinical settings, dance programs for people living with Parkinson's have been implemented in communities worldwide, yet more research is needed to characterize their effects and understand their broader impact. This thesis aims to assess the effects and impact of dance using mixed methods. First, a systematic review of randomized controlled trials was conducted to assess the efficacy of dance for people living with Parkinson's and the quality of existing evidence. Then, the feasibility of implementing a dance program for people living with Parkinson's and conducting an intervention trial in Northern Ireland was assessed and preliminary analyses were conducted on dance's effects on motor and non-motor outcomes. Based on feedback from participants in our feasibility study and in the literature that dancing can have an immediate, short-term effect on mobility, mood and energy, a follow up study was conducted to investigate the short-term effects of dance on several motor and non-motor outcomes. The final participatory study explored the lived experience of dancing in people living with Parkinson's and Multiple Sclerosis and its impact on the body, self and relationships with others using a range of qualitative methods, including semi-structured interviews and observations analyzed using a phenomenological approach. This thesis concludes that dance has the potential to impact movement of the body, one's state of mind, relational health, and the quality and movement of life in people living with Parkinson's. Many people living with Parkinson's and other chronic illnesses who dance seem to make it a part of their lives due to its holistic impact on several dimensions of health and the passion that develops from the practice.

Successful navigation through our environment depends on selecting and controlling the appropriate course of action. According to ecological psychology, decisions on when and how we act is achieved through the detection of information specifying affordances; the fit between an actor’s action possibilities and their environment. As a result, errors in decision-making may derive from an unanticipated actualization of what the environment affords. While recent research has aimed to understand how changes in physical limitations of individuals influence their perception of what the environment affords them, little research has investigated how changes during life span influence sensitivity to specifying information, and how these may impact upon action decisions. In a series of interlinking studies using virtual traffic environments, different age groups were investigated in their ability to: tune into action-relevant information in the optical array, use such information to guide both the choice of behaviour and the continuous guidance of action, and re-educate attention to specifying or non-specifying action information. Findings support the general assumption that ageing impacts the ability to tune into information which specifies whether a gap affords crossing or not and this extends beyond action selection to influence how an affordance is actualized. Together, these studies provide evidence of age deterioration in the use of perceptual-based information that guides decisions about when and how to act. This suggests that to help prevent unintended injury, intervention strategies for older adults should focus on re-learning how to tune into and use information that specifies affordances in hazardous environments.

An important goal in evolutionary biology is to conceptually incorporate phenotypic plasticity into the framework of trait evolution. Here, I used a system of extremophile fish, Poecilia mexicana, locally adapted to all combinations of sulfidic and cave habitats, to test for genetic and plastic variation in energy metabolism and potential links between physiological and behavioral traits. Energy metabolism was predicted by habitat of origin, with cavefish showing higher routine metabolic rates, and sulfidic fish generally showing higher peak metabolic rates, although the extent to which was dependent on resource availability. Together variation in routine and peak metabolic rate lead to complex variation in metabolic scope, which varied depending on the presence of light and sulfide in natural populations as well as resource availability. A combination of resource availability, population of origin, and metabolic physiology (metabolic scope) also predicted male mate choice behavior, which was measured in two different ways: strength of preference for a preferred mate and the number of switches between two potential mates. Strength of preference depended on metabolic scope and the presence of sulfide in the population of origin, with sulfidic fish showing consistently higher strengths of preference, and strength of preference decreasing with metabolic scope overall. The number of switches between stimuli depended on presence of light in the population of origin, food treatment, and metabolic scope. Well-fed P. mexicana from cave populations increased their switching behavior with their metabolic scope. These results are part of a larger framework that elucidates the complex interaction between energetics, behavior, their genetic underpinnings, and aspects of the environment that affect the two phenotypes and the relationships between them.

In cognitive neuroscience and neuropsychology, the collection of cognitive and behavioural data is predominantly achieved by implementing paper-and-pencil and computerized (i.e., 2D and 3D applications) assessments. However, these psychometric tools in clinics and/or laboratories display several limitations and discrepancies between the observed performance in the laboratory/clinic and the actual performance of individuals in everyday life. The functional and predictive association between an individual's performance on a set of neuropsychological tests and the individual's performance in various everyday life settings is called ecological validity. Ecological validity is considered an important issue that cannot be resolved by the currently available assessment tools. Virtual reality head-mounted displays (HMD) appear to be effective research tools, which may address the problem of ecological validity in neuropsychological testing. However, their widespread implementation is hindered by virtual reality induced symptoms and effects (VRISE) and the lack of skills in virtual reality software development. In this PhD, a technological systematic literature review of the reasons for adverse symptomatology was conducted and suggestions and technological knowledge for the implementation of virtual reality HMD systems in cognitive neuroscience provided. The review indicated features pertinent to display, sound, motion tracking, navigation, ergonomic interactions, user experience, and computer hardware that should be considered by researchers. Subsequently, a meta-analysis of 44 neuroscientific or neuropsychological studies involving virtual reality HMD systems was performed. The meta-analysis of the virtual reality studies demonstrated that new generation HMDs induce significantly less VRISE and marginally fewer dropouts. Importantly, the commercial versions of the new generation HMDs with ergonomic interactions had zero incidents of adverse symptomatology and dropouts. HMDs equivalent to or greater than the commercial versions of contemporary HMDs accompanied with ergonomic interactions are suitable for implementation in cognitive neuroscience. Another aim of this PhD was to devise a brief tool to appraise and report both the quality of software features and VRISE intensity quantitatively; such a tool does not currently exist. The Virtual Reality Neuroscience Questionnaire (VRNQ; Kourtesis et al., 2019) was developed to assess the quality of virtual reality software in terms of user experience, game mechanics, in-game assistance, and VRISE. Forty participants aged between 28 and 43 years were recruited (18 gamers and 22 non-gamers) for the study. They participated in 3 different virtual reality sessions until they felt weary or discomfort and subsequently filled in the VRNQ. The results demonstrated that VRNQ is a valid tool for assessing virtual reality software as it has good convergent, discriminant, and construct validity. The maximum duration of virtual reality sessions should be between 55 and 70 min when the virtual reality software meets or exceeds the parsimonious cut-offs of the VRNQ, and the users are familiarized with the virtual reality system. Also, gaming experience does not affect how long virtual reality sessions should last. Furthermore, while the quality of virtual reality software substantially modulates the maximum duration of virtual reality sessions, age and education do not. Finally, deeper immersion, better quality of graphics and sound, and more helpful in-game instructions and prompts were found to reduce VRISE intensity. The VRNQ facilitates the brief assessment and reporting of the quality of virtual reality software features and/or the intensity of VRISE, while its minimum and parsimonious cut-offs may appraise the suitability of virtual reality software for implementation in research and clinical settings. However, the development of virtual reality software is predominantly dependent on third parties (e.g., freelancers or companies) with programming and software development skills. A solution that will promote the adoption of immersive virtual reality as a research and clinical tool might be the in-house development of virtual reality research/clinical software by computer science literate cognitive scientists or research software engineers. In Chapter 4, guidelines are offered for the development of virtual reality software in cognitive neuroscience and neuropsychology, by describing and discussing the stages of the development of Virtual Reality Everyday Assessment Lab (VR-EAL), the first neuropsychological battery in immersive virtual reality. Techniques for evaluating cognitive functions within a realistic storyline are discussed. The utility of various assets in Unity, software development kits, and other software are described so that cognitive scientists can overcome challenges pertinent to VRISE and the quality of the virtual reality software. In addition, VR-EAL is evaluated in accordance with the necessary criteria for virtual reality software for research purposes. The virtual reality neuroscience questionnaire (VRNQ) was implemented to appraise the quality of the three versions of VR-EAL in terms of user experience, game mechanics, in-game assistance, and VRISE. Twenty-five participants aged between 20 and 45 years with 12-16 years of full-time education evaluated various versions of VR-EAL. The final version of VR-EAL achieved high scores in every sub-score of the VRNQ and exceeded its parsimonious cut-offs. It also appeared to have better in-game assistance and game mechanics, while its improved graphics substantially increased the quality of the user experience and almost eradicated VRISE. The results substantially support the feasibility of the development of effective virtual reality research and clinical software without the presence of VRISE during a 60-min virtual reality session. In Chapter 5, validation of VR-EAL as an assessment of prospective memory, episodic memory, attention, and executive functions using an ecologically valid approach is examined. Performance on the VR-EAL, an immersive virtual reality neuropsychological battery, is examined against an extensive paper-and-pencil neuropsychological battery. Forty-one participants (21 females) were recruited: 18 gamers and 23 non-gamers who attended both an immersive virtual reality and a paper-and-pencil testing session. Bayesian Pearson correlation analyses were conducted to assess construct and convergent validity of the VR-EAL. Bayesian t-tests were performed to compare virtual reality and paper-and-pencil testing in terms of administration time, similarity to real life tasks (i.e., ecological validity), and pleasantness. VR-EAL scores were significantly correlated with their equivalent scores on the paper-and-pencil tests. The participants' reports indicated that the VR-EAL tasks were considered significantly more ecologically valid and pleasant than the paper-and-pencil neuropsychological battery. The VR-EAL battery also had a shorter administration time. The VR-EAL appears to be an effective neuropsychological tool for the assessment of everyday cognitive functions, and has enhanced ecological validity, a highly pleasant testing experience, and does not induce cybersickness. In the final part of this thesis, the preparatory attentional and memory (PAM) and the multiprocess theories of prospective memory are examined by attempting to identify the cognitive functions which may predict the individual's performance on ecologically valid prospective memory tasks in the same group of participants described in Chapter 5. Bayesian t-tests were conducted to explore the differences among different prospective memory tasks (e.g., event-based and time-based) and prospective memory tasks with varying delays between encoding and the recall of the intended action (e.g., short-delay versus long-delay). Bayesian linear regression analyses were performed to examine the predictors of VR-EAL scores. The results revealed that the type of prospective memory task does not play a significant role in everyday prospective memory functioning, but instead the length of delay between encoding and retrieving the prospective memory intention plays a central role. Support for the PAM and MP frameworks was found in non-focal and focal event-based tasks respectively. However, the findings, inferring a dynamic interplay between automatic and intentional monitoring and retrieval processes, agree with the inclusive approach of the multiprocess framework. Also, the role of executive functions appears crucial in everyday PM. Finally, everyday PM is predominantly facilitated by episodic memory, visuospatial attention, auditory attention, and executive functions. In conclusion, this PhD thesis attempted to show how immersive virtual reality research methods may be implemented efficiently without the confounding effect of cybersickness symptomatology in order to enhance the ecological validity of neuropsychological testing and contribute to our understanding of everyday cognitive ability.

It was the purpose of this investigation to study the effect of varying the direct to reverse time ratio of periodically reversed direct current from 1.0 to 20.0 on the electrolytic production of lead chromate at an average anode current density of 0.0059 amperes per square centimeter, and to study the effect of 60 and 502.3 ± 7.7 cycles per second from 0.00113 to 0.01546 amperes per square centimeter of peak superimposed alternating current on direct current on the yield of lead chromate prepared by the electrolysis of a bath containing potassium chromate and sodium nitrate between lead electrodes. Electrolysis of a bath containing 3.60 grams of potassium chromate, 11.62 grams of sodium nitrate, end 1000 grams of water with simple direct current yielded 6.07 grams of lead chromate per ampere-hour, the purity being 92.7 percent lead chromate. The anode current density was 0.0049 amperes per square centimeter, and the current efficiency was 98.1 percent. During the electrolysis, by maintaining the ph of the electrolyte at 6.0, by adding a solution to 2.0 weight percent chromic acid, the purity of the product was increased. Electrolysis of the same bath using periodically reversed direct current yielded 4.53 grams of lead chromate per ampere-hour, the purity being 93.9 percent lead chromate. The time ratio was 20.0, the anode current density was 0.0049 amperes per square centimeter, and the current efficiency was 66.75 percent. Decreasing the direct to reverse time ratio gave lower yields and purities. Apparently, there is no advantage in using periodically reversed direct current over the use of direct current for this reaction under the above conditions. Electrolysis of the same bath with alternating current superimposed on direct current yielded 5.49 grams of lead chromate per direct current ampere-hour, at a purity of 99.4 percent lead chromate, when using 494.7 cycle alternating current. The alternating and direct current densities were 0.0078 and 0.0048 amperes per square centimeter, respectively. This was the purest product obtained in this investigation. Use of 60 cycle alternating current yielded 3.83 grams of lead chromate per direct current ampere-hour, at a purity of 93.9 percent lead chromate. The alternating and direct current densities were 0.00141 and 0.00484 amperes per square centimeter. Further experiments were made using direct current and periodically reversed direct current on a bath containing 6.80 grams of potassium chromate, 8.14 grams of sodium nitrate, and 1000 grams of water. Results from these electrolytes were much poorer than those obtained with the previous bath, so no experiments with superimposed alternating current on direct current were carried out with this latter bath.

Massive nylon lends itself to many applications because of its outstanding toughness, abrasion-resistance, and chemical resistance. These properties make it a satisfactory bearing material, and in some applications it is superior to metallic bearings. However, the disadvantages of low melting point and low thermal conductivity prevent nylon from being used more extensively. A previous investigation of the effect on thermal conductivity of change in crystallinity brough about by rolling and annealing nylon 66 indicated that it exhibits anisotropy of conductivity and that its thermal conductivity depends on primary physical factors related to its molecular structure. The purpose of this investigation was to determine the relationships of anisotropy of thermal conductivity, bond strength, degree of crystallinity, molecular orientation, and the changes resulting from the rolling-annealing treatment of nylon 66. A survey was made of the literature on the thermal conductivity of nylon in particular and non-metallic solids in general, of anisotropy of thermal conductivity, of the internal structure of nylon, of the effect of physical treatment on the internal structure and physical properties of nylon, and on test methods. The experimental part of the work consisted of measuring the thermal conductivity of nylon along three directions of heat flux, and the changes in internal structure and thermal conductivity in the three directions of heat flux resulting from rolling and annealing treatment. The thermal conductivity of six samples of nylon 66 was determined. The first three samples were cut from a slab of cast nylon ¼ inch thick. The conductivity was measured in a direction perpendicular to the plane of the sample and to the greatest length of the slab, in the plane of the sample, but perpendicular to the greatest length of the slab, and in the plane and in the direction of the greatest length of the slab. The other three samples were prepared from a ¼-inch thick cast slab which had been cold-rolled to half the original thickness and then annealed for two hours at 240 °C. To determine the effect of physical treatment on the internal structure and thermal conductivity, the density and degree of crystallinity were determined and x-ray diffraction patterns of the samples were made. Standard laboratory procedures were used in all of the tests. The thermal conductivity was measured using the ASTM method Cl77-45, the guarded hot plate method. In this method two sheets of nylon five inches square were sandwiched between an electrical heater and two brass cooling blocks. The quantity of heat which flowed through the samples was measured by measuring the electric power input to the heater. The temperature drop across the samples was measured by means of thermocouples. The sample thickness and area were measured, and from these quantities the thermal conductivity was calculated. The density was determined by weighing one to six-gram samples in air and in water, according to the standard method. The degree of crystallinity was calculated from the density of dried samples by assuming a linear relationship between density and degree of crystallinity. This method of calculation was worked out by Hermans for cellulose and used by Snow for nylon. The results of this investigation showed that when nylon slabs were rolled and annealed, the degree of crystallinity was increased, confirming the conclusions of previous investigators. The thermal conductivity was affected mainly by an orientation of the molecule produced by rolling rather than by the change in degree of crystallinity. Because of a preferential orientation of the molecules in the direction of rolling, the nylon thermal conductivity became anisotropic. These results support Rehner’s hypothesis that the thermal conductivity of polymers is mainly dependent on the strength of the bonds in the molecular chain, and that thermal conduction in polymers is mainly molecular conduction, not lattice conduction. The following conclusions were reached from tests made on the nylon slabs.