abstract: In investigating mediating processes, researchers usually use randomized experiments and linear regression or structural equation modeling to determine if the treatment affects the hypothesized mediator and if the mediator affects the targeted outcome. However, randomizing the treatment will not yield accurate causal path estimates unless certain assumptions are satisfied. Since randomization of the mediator may not be plausible for most studies (i.e., the mediator status is not randomly assigned, but self-selected by participants), both the direct and indirect effects may be biased by confounding variables. The purpose of this dissertation is (1) to investigate the extent to which traditional mediation methods are affected by confounding variables and (2) to assess the statistical performance of several modern methods to address confounding variable effects in mediation analysis. This dissertation first reviewed the theoretical foundations of causal inference in statistical mediation analysis, modern statistical analysis for causal inference, and then described different methods to estimate causal direct and indirect effects in the presence of two post-treatment confounders. A large simulation study was designed to evaluate the extent to which ordinary regression and modern causal inference methods are able to obtain correct estimates of the direct and indirect effects when confounding variables that are present in the population are not included in the analysis. Five methods were compared in terms of bias, relative bias, mean square error, statistical power, Type I error rates, and confidence interval coverage to test how robust the methods are to the violation of the no unmeasured confounders assumption and confounder effect sizes. The methods explored were linear regression with adjustment, inverse propensity weighting, inverse propensity weighting with truncated weights, sequential g-estimation, and a doubly robust sequential g-estimation. Results showed that in estimating the direct and indirect effects, in general, sequential g-estimation performed the best in terms of bias, Type I error rates, power, and coverage across different confounder effect, direct effect, and sample sizes when all confounders were included in the estimation. When one of the two confounders were omitted from the estimation process, in general, none of the methods had acceptable relative bias in the simulation study. Omitting one of the confounders from estimation corresponded to the common case in mediation studies where no measure of a confounder is available but a confounder may affect the analysis. Failing to measure potential post-treatment confounder variables in a mediation model leads to biased estimates regardless of the analysis method used and emphasizes the importance of sensitivity analysis for causal mediation analysis. / Dissertation/Thesis / Ph.D. Psychology 2013

Orientador: Carlos Alberto dos Reis Filho / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-10T09:10:45Z (GMT). No. of bitstreams: 1 Nadalin_EvertonZaccaria_M.pdf: 2191420 bytes, checksum: 38decb6d182261bd5e5fc2df03ad1b9a (MD5) Previous issue date: 2007 / Resumo: Este trabalho está relacionado com o problema da pesagem de um bovino criado em campo aberto, não confinado, destinado ao corte. O conhecimento da evolução contínua do peso do animal é de grande importância neste ramo de negócios e uma solução tecnicamente satisfatória, de custo aceitável e de tecnologia proprietária, certamente representa uma significativa contribuição ao Brasil. Uma solução idealizada, que é a meta maior do projeto em que este trabalho se enquadra, prevê o uso de um dispositivo instalado nas patas do animal, contendo um sensor de pressão ou de impacto e uma unidade eletrônica local que condiciona, digitaliza e armazena os sinais do sensor. Além disto, a unidade eletrônica transmite estes dados através de uma rede de comunicação de dados sem fio e de baixo consumo de energia para centrais de coleta e processamento estrategicamente situadas na área em que o gado se desenvolve. Tal sistema viabiliza, deste modo, acumular no tempo as informações de cada pisada de cada animal. O sinal produzido por um sensor em conseqüência do impacto de cada pisada é influenciado por diversos parâmetros, dentre os quais o peso do animal. Sendo assim, é impossível estabelecer uma relação direta e exclusiva entre o sinal do sensor e o peso. Entretanto, os valores acumulados destes sinais descrevem uma função de múltiplas variáveis que pode ser tratada juntamente com algumas informações complementares sobre o ambiente, buscando-se o isolamento da variável peso. Para isto, uma abordagem adequada prevê o desenvolvimento ou adoção de um modelo do mecanismo de movimentação do animal e o tratamento numérico dos dados acumulados aplicando filtragens e buscas de correlações. Trata-se, portanto, de uma solução de natureza multidisciplinar, que exige uma aliança de conhecimentos complementares para a sua realização. Coube ao presente trabalho, desenvolvido no âmbito da engenharia eletrônica, a implementação de uma etapa preliminar e necessária que consiste de um sistema de caracterização de pisadas constituído de um conjunto de sensores piezoelétricos, uma interface que permite a transferência dos dados destes sensores a um computador e um ambiente de software através do qual foram testados algoritmos para a verificação de correlação entre os sinais associados às pisadas e o peso do agente. São frutos deste trabalho um sistema de coleta automática dos sinais de pisadas, que permitiu a captura de sinais com sensores instalados tanto numa plataforma fixa como num sapato tênis, e um estudo comparativo dos resultados obtidos de tratamentos numéricos distintos aplicados aos dados experimentais / Abstract: This work deals with the problem of weighting non-confined cattle raised in open field for meat production. The knowledge of the continuous weight change of the animal is of great importance in this business field. Therefore, a technically satisfactory solution, with acceptable price and customized technology, certainly represents a significant contribution to Brazil. The idealized solution, which is the major goal to be pursued by the project in which this work takes part, considers the use of a device placed into the animal¿s hooves. It would contain a pressure or impact sensor and a local electronic unit, which conditions, digitizes and stores the signals received from the sensor. The electronic unit also transmits this data through a low-energy wireless communication network to reception and processing stations placed strategically around the area where the cattle raises. Such a system enables accumulating data about the steps of each animal throughout time. Several parameters, including the animal¿s weight, influence the signal produced by a sensor as a consequence to the impact of a step. Therefore, it is impossible to establish a direct and exclusive relation between the signal given by the sensor and the weight. Nevertheless, the accumulated data of these signals describe a multiple variable function that can be treated together with some complementary information about the environment, seeking the isolation of the weight variable. An adequate approach predicts the development or use of a model of the animal¿s locomotion mechanism and the numeric treatment of the accumulated data by applying filtering and seeking correlations. The multidisciplinary nature of this solution demands an alliance of complementary knowledge for its accomplishment. This work, developed in the scope of electrical engineering, implemented a necessary and preliminary stage that consists of: a step characterization system, made of a set of piezoelectric sensors, an interface that allows transferring data from the sensors to a computer and a software environment, in which algorithms were tested to verify the correlation between the signals associated to the steps and the agent¿s weight. The main result of this work are a system that performs automatic collecting of steps, which allowed the capture of the signals with sensors installed both on a fixed platform and inside tennis shoes, and the comparative study of the results of different numerical treatments applied to the experimental data / Mestrado / Eletrônica, Microeletrônica e Optoeletrônica / Mestre em Engenharia Elétrica

Tomadas de decisão intertemporais são influenciadas não somente pelo efeito de desconto do valor em diferentes instantes, como também pelo efeito de percepção temporal. Uma das principais dificuldades que afetam experimentos padrões envolvendo estas escolhas é a simultaneidade de ambos os efeitos no processo de desconto. Através da unificação das leis da psicofísica de percepção do atraso e da associação destas às funções de desconto de valor, propusemos uma forma generalizada para o processo de desconto intertemporal envolvendo ambos os domínios. Mostramos também que a propriedade aditiva do grau de inconsistência, grandeza obtida a partir das funções de desconto, permite discriminar a influência de cada efeito no processo de tomada de decisão. De forma similar ao proposto para escolhas intertemporais, estendemos a teoria psicofísica de percepção ao domínio das probabilidades. Adotando a perspectiva de que o atraso médio para recebimento de uma recompensa está relacionado à sua probabilidade de recebimento, obtivemos uma função de desconto probabilística que abrange os efeitos de desconto de valor e de percepção de probabilidades. Em paralelo ao desenvolvimento no domínio experimental, exploramos também os modelos teóricos (axiomáticos) que fundamentam as tomadas de decisão probabilísticas. Propusemos que a forma da função peso de probabilidades, exploradas em modelos como a teoria da utilidade esperada dependente da ordenação e a teoria do prospecto acumulado, seja representada pela função de desconto generalizada que obtivemos a partir dos modelos fenomenológicos. Neste caso, a função peso é amparada por modelos fenomenológicos de decisão e deriva da suposição de que indivíduos se comportam de forma similar frente a probabilidades e atrasos. / Intertemporal decision making are influenced not only by the effect of the reward value discount at different moments, but also by the time perception effect. One of the main difficulties that affects standard experiments involving intertemporal choices is the simultaneity of both effects on the discount. Through the unification of the psychophysical laws of delay perception and their association to the value discounting functions, we proposed a generalized process for the intertemporal discount involving both domains. We also showed that the additive property of the inconsistency degree, quantity obtained from the discount functions, allowed us to distinguish between both effects in decision making. As proposed for intertemporal choices, we extended the psychophysical perception theory to the probabilities domain. Adopting the perspective that the average delay for receiving a reward is related to the probability of receiving it, we obtained a probabilistic discount function covering value discounting and probability perception. Parallel to the experimental development, we also explored theoretical models that underlie probabilistic decision making. We proposed that the shape of the probability weighting function, explored in models such as rank-dependent expected utility theory and cumulative prospect theory, should be described by the generalized probabilistic function that we obtained from the phenomenological models. Therefore, the weighting function is supported by phenomenological models and derives from the assumption that subjects behave similarly in front of probabilities and delays.

Medidas de transmitância em óculos de sol fazem parte de alguns dos testes propostos pela norma NBR15111, porém, observou-se que não há no país um sistema preciso para medidas de transmitância que não seja com o uso de espectrofotômetros e pessoas qualificadas para utilizar o equipamento, aplicar as funções de ponderação e fornecer laudos técnicos sobre a lente testada. Assim, o trabalho propõe métodos que utilize componentes simples e resulte na construção de um equipamento que possa fornecer a informação sobre as transmitâncias dos filtros de óculos de sol de forma automática, de acordo com a NBR15111. Foi feita a espectrofotometria de 45 lentes de óculos de uso geral, e calculadas suas respectivas transmitâncias (luminosa, ultravioleta, semafórica), como amostras de comparação com as metodologias testadas. Os testes incluíram lâmpadas e LEDs como fontes de emissão para a região do espectro visível e lâmpadas fluorescentes para a região ultravioleta, além de sensores específicos para cada região espectral, e eletrônica de controle e aquisição dos sinais. Foram usadas ferramentas de análise estatística, como erro RMS, coeficiente de aderência (GFC) e análise de Bland-Altman. Aproximadamente 62% dos óculos testados foram reprovados para a direção no trânsito e nenhum foi reprovado para a proteção UV. Os métodos RMS e GFC apontaram combinações entre fonte e sensor que conseguem fazer a ponderação com os menores desvios-padrão das medidas espectrométricas. O sistema para teste UV mostrou melhoria significativa em relação aos desenvolvidos anteriormente, com um desvio padrão de 4,45 pontos percentuais e o sistema para testes na região visível utilizou o sensor TCS, no qual a regressão polinomial mostrou melhores resultados que as redes neurais artificiais, com os menores limites de concordância para as transmitâncias. Um número maior de amostras pode melhorar os métodos para que se obtenha uma aproximação ideal para todos os óculos de sol. Uma montagem com LEDs se apresenta como uma alternativa interessante quanto ao custo, tamanho e baixo consumo para medidores de transmitância portáteis. Desenvolveu-se um sistema com módulos para teste UV, Visível e Semafórico, para uso público no campus da USP de São Carlos, proporcionando conscientização e extensão dos estudos em relação aos óculos solares, contribuindo efetivamente com a população brasileira. / Transmittance measurements in sunglasses are a part of the standard tests proposed by NBR15111, however, it was observed that there is no accurate system for measuring transmittance different than using spectrophotometers and skilled practitioner to manage the equipment, apply the weighting functions and provide technical reports on the lens tested. Therefore, this study proposes methodologies using simple components and that can result a device that can provide information about sunglass filters automatically, according to NBR15111. Spectrophotometry was calculated in 45 lenses of sunglasses for general use, and their respective transmittance (visible, ultraviolet, traffic light), used as samples for testing the methodologies. The tests included lamps and LEDs as light sources for the visible region spectrum and fluorescent lamps for ultraviolet region, specific sensors for each spectral region, and electronic for control and signal acquisition. Statistical analysis tools were used, such as RMS error, goodness-of-fit coefficient (GFC) and the Bland - Altman analysis. Approximately 62% of the tested lenses failed in the test for signal light recognition and none has failed for UV protection. The RMS and GFC methods indicated combinations between source and sensor that calculated transmittance with the lowest standard deviations of the spectrometric measurements. The system for testing UV showed improvement compared to other previously equipment, with a standard deviation of 4.45 and the system for testing the visible region used the TCS sensor, in which the polynomial regression showed better results than artificial neural networks. A larger number of samples can improve the methods in order to obtain an optimal calibration that includes all sunglasses. LEDs seem to be a good alternative in terms of cost, size and low power consumption for portable transmittance meters. We developed a system with modules for testing UV, Visible and Traffic Light transmittance for public use on the USP São Carlos campus, providing awareness and extent of studies about the sunglasses, in order to contribute more effectively to the Brazilian population.

L’objectif du présent travail de thèse est d’étudier les mécanismes de pondération des informations visuelles et somatosensorielles dans le contrôle du mouvement volontaire visuoguidé de la main lorsque les retours sensoriels de ces deux canaux véhiculent des informations spatiales congruentes ou incongruentes. Une incongruence entre les informations visuelles et somatosensorielles peut-être crée expérimentalement en décalant l’environnement visuel perçu des participants. Dans une telle situation, les participants devaient suivre les contours d’une forme géométrique irrégulière avec un stylet sur une tablette graphique. L’activité cérébrale des régions visuelles, somatosensorielles et pariétales postérieures a été enregistrée en électroencéphalographie, et quantifiée par la mesure de l’amplitude de potentiels évoqués visuels [Etude 1] et de la puissance des bandes de fréquences alpha (8-12 Hz), beta (15-25 Hz) et gamma (50-80 Hz) [Etudes 2 et 3]. Nous avons ainsi montré que le mouvement visuoguidé entraînait une augmentation de l’activité au niveau des aires corticales visuelles, et que le contrôle du mouvement en situation d’incongruence induisait une augmentation supplémentaire de l’excitabilité des cortex visuels, somatosensoriels et pariétaux postérieurs. Ces modulations reflèteraient des mécanismes de pondération du traitement de ces entrées sensorielles dans le but de s’adapter à cette situation. Plus généralement, nos résultats soutiennent l’idée que notre système nerveux est en mesure de moduler localement son activité en fonction de la pertinence du traitement des informations pour répondre aux exigences imposées par le contexte. / The goal of our doctoral research was to investigate the weighting of these sensory inputs in conditions under which they provided either congruent or incongruent information about hand motion. A visuo-somatosensory incongruence can be induced experimentally by shifting the visual feedback of the environment. We asked participants to follow precisely the outline of an irregular shape with a stylus on a digitizing tablet. Brain activity was recorded with an electroencephalographic device, and quantified by measuring visual evoked potentials amplitudes [Study 1], and the power in the alpha (8-12 Hz), beta (15- 25 Hz) and gamma (50-80 Hz) frequency-bands [Studies 2 and 3]. We first evidenced that visually-guided hand movements increased the sensitivity to visual inputs of a large cortical network. Moreover, we showed that controlling movement in a situation with an incongruence between visual and somatosensory input led to a further increase of visual, somatosensory and posterior-parietal cortical excitability. We suggest that these modulations reflect sensory weighting mechanisms in order to attempt to adapt to the sensory incongruence. Interestingly, in the somatosensory areas, we found that the sensory incongruent condition led to a reduction of gamma power, suggesting a reduced integration of somatosensory inputs for controlling movements. Taken together, our findings are in line with the existence of a general sensory gain control mechanism driven by the state of adaptation of the sensorimotor system in a given sensory context. More generally, our results argue for the idea that sensory processing is function of the context-dependent relevance of the sensory inputs.

When using visual and proprioceptive information to plan a reach, it has been proposed that the brain combines these cues to estimate the object and/or limb’s location. Specifically, according to the maximum-likelihood estimation (MLE) model, more reliable sensory inputs are assigned a greater weight (Ernst & Banks, 2002). In this research we examined if the brain is able to adjust which sensory cue it weights the most. Specifically, we asked if the brain changes how it weights sensory information when the availability of a visual cue is manipulated. Twenty-four healthy subjects reached to visual (V), proprioceptive (P), or visual + proprioceptive (VP) targets under different visual delay conditions (e.g. on V and VP trials, the visual target was available for the entire reach, it was removed with the go-signal or it was removed 1, 2 or 5 seconds before the go-signal). Subjects completed 5 blocks of trials, with 90 trials per block. For 12 subjects, the visual delay was kept consistent within a block of trials, while for the other 12 subjects, different visual delays were intermixed within a block of trials. To establish which sensory cue subjects weighted the most, we compared endpoint positions achieved on V and P reaches to VP reaches. Results indicated that all subjects weighted sensory cues in accordance with the MLE model across all delay conditions and that these weights were similar regardless of the visual delay. Moreover, while errors increased with longer visual delays, there was no change in reaching variance. Thus, manipulating the visual environment was not enough to change subjects’ weighting strategy, further i

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第22828号 / 医博第4667号 / 新制||医||1047(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 村井 俊哉, 教授 古川 壽亮, 教授 高橋 淳 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

No description available.

The diploma thesis is a part of the project Intelligent weighting and sorting of goods. The principle of measuring of weight from movement dynamics with derivation calculation relation is detailed described in the first part. The description of used automation means is also in the first part of the diploma thesis. The second part includes design of three adaptive control systems for controlling DC electromotor’s arm, communication principles between programmable logic controllers and designed code of program.

In underground construction and foundation engineering uncertainties associated with subsurface properties are inevitable to deal with. Site investigations are expensive to perform, but a limited understanding of the subsurface may result in major problems; which often lead to an unexpected increase in the overall cost of the construction project. This study aims to optimize the pre-investigation program to get as much correct information out from a limited input of resources, thus making it as cost effective as possible. To optimize site investigation using soil-rock sounding three different sampling techniques, a varying number of sample points and two different interpolation methods (Inverse distance weighting and point Kriging) were tested on four modeled reference surfaces. The accuracy of rock surface predictions was evaluated using a 3D gridding and modeling computer software (Surfer 8.02®). Samples with continuously distributed data, resembling profile lines from geophysical surveys were used to evaluate how this could improve the accuracy of the prediction compared to adding additional sampling points. The study explains the correlation between the number of sampling points and the accuracy of the prediction obtained using different interpolators. Most importantly it shows how continuous data significantly improves the accuracy of the rock surface predictions and therefore concludes that geophysical measurement should be used combined with traditional soil rock sounding to optimize the pre-investigation program.