Neste trabalho, tratamos de uma imagem que circula em nosso país sobre o espanhol e sua aprendizagem, segundo a qual os chamados \"falsos amigos\" constituem as principais dificuldades para o brasileiro que deseja aprender essa língua estrangeira. Partimos de uma hipótese de Celada & González (2000) de que tal modo de interpretação desse processo foi inaugurado por Nascentes (1939) quem, por sua vez, recolheu em seus estudos muitas das impressões do senso comum de sua época a esse respeito. Tal tradição norteou e em grande parte norteia ainda muitos trabalhos científicos no Brasil a respeito do ensino e da aprendizagem do espanhol, os quais se baseiam, como a interpretação tradicional que lhes dá suporte, nos pressupostos teóricos do modelo da Análise Contrastiva em sua versão forte. Primeiramente fazemos uma análise dos termos mais comuns em circulação (falsos cognatos, heterosemánticos, falsos amigos), mostrando que não fazem necessariamente referência a fenômenos da mesma natureza, e posteriormente questionamos a idéia corrente de que os falsos amigos seriam o grande vilão da aprendizagem do espanhol por brasileiros. Nossa proposta nesta pesquisa é tratar a aprendizagem do espanhol por brasileiros e a formação de sentido de outra perspectiva teórica, que nos permita observar fatores que oferecem restrições às línguas e que atuam na produção dos efeitos de sentido no contato desses dois idiomas; esses fatores serão, essencialmente, a história e as representações sociais. / The present investigation refers to the traditional perception that exists in Brazil about the Spanish language and its learning by Brazilians. In such perception, the uttermost obstacle that any Brazilian must overcome in his process of learning the Spanish language are the denominated \"false friends\". According to Celada & González (2000), such perception has been introduced in Brazil by Nascentes (1939), in whose work one is able to find most of the common sense and common lore opinions about the learning of the Spanish language by Brazilians. Such perception has ever since dominated most of the Brazilian academic investigation about the teaching and learning of the Spanish language, which theoretic basis can be found in the \"strong\" version of the Contrastive Analysis. This investigation begins with the analysis of the terms most commonly used by the academic investigation (such as false friends, false cognates, heterosemánticos), in order to demonstrate that such terms do not refer to the same phenomena. In the sequence, the perception that the \"false friends\" are the greatest villain in the learning of the Spanish language by Brazilians is put in doubt. The final objective of this investigation is to study the process of learning of the Spanish language by Brazilians, and from another theoretic perspective, to analyse the production of meaning that allows to determine the factors (specially history and social representation) which restrict the languages, and influence the production of the meaning in the contact between Brazilian Portuguese and Spanish.

Maritime surveillance radars are critical in commerce, transportation, navigation, and defense. However, the sea environment is perhaps the most challenging of natural radar backdrops because maritime radars must contend with electromagnetic backscatter from the sea surface, or sea clutter. Sea clutter poses unique challenges in very low grazing angle geometries, where typical statistical assumptions regarding sea clutter backscatter do not hold. As a result, traditional constant false alarm rate (CFAR) detection schemes may yield a large number of false alarms while objects of interest may be challenging to detect. Solutions posed in the literature to date have been either computationally impractical or lacked robustness. This dissertation explores whether fully polarimetric radar offers a means of enhancing detection performance in low grazing angle sea clutter. To this end, MIT Lincoln Laboratory funded an experimental data collection using a fully polarimetric X-band radar assembled largely from commercial off-the-shelf components. The Point de Chene Dataset, collected on the Atlantic coast of Massachusetts’ Cape Ann in October 2015, comprises multiple sea states, bandwidths, and various objects of opportunity. The dataset also comprises three different polarimetric transmit schemes. In addition to discussing the radar, the dataset, and associated post-processing, this dissertation presents a derivation showing that an established multiple input, multiple output radar technique provides a novel means of simultaneous polarimetric scattering matrix measurement. A novel scheme for polarimetric radar calibration using a single active calibration target is also presented. Subsequent research leveraged this dataset to develop Polarimetric Co-location Layering (PCL), a practical algorithm for mitigation of low grazing angle sea clutter, which is the most significant contribution of this dissertation. PCL routinely achieves a significant reduction in the standard CFAR false alarm rate while maintaining detections on objects of interest. Moreover, PCL is elegant: It exploits fundamental characteristics of both sea clutter and object returns to determine which CFAR detections are due to sea clutter. We demonstrate that PCL is robust across a range of bandwidths, pulse repetition frequencies, and object types. Finally, we show that PCL integrates in parallel into the standard radar signal processing chain without incurring a computational time penalty.

L’exocytose vésiculaire est une voie physiologique majeure de la communication intercellulaire. Dans ce contexte, le TIRFM (microscopie de fluorescence par réflexion totale interne) et l’ampérométrie sont aujourd'hui les deux méthodes analytiques les plus fréquemment utilisées dans l’étude de l’exocytose. En raison de la complémentarité de ces deux techniques d’analyse pour le suivi de la sécrétion exocytotique, leur combinaison pour suivre la sécrétion exocytotique a d'abord été réalisée par notre groupe en 2011. Ce couplage a permis un enregistrement simultané des signaux fluorescents et ampérométriques avec une bonne résolution spatiale et temporelle. L'inconvénient majeur de ce travail reste le chargement indépendant des sondes optique et électrochimique dans les vésicules de sécrétion, ce qui entraîne la détection d’évènements « orphelins » ampérométriques ou optiques ainsi que la faible efficacité de détection des évènements couplés. Par conséquent, dans cette thèse, nous avons tenté de mettre à profit une sonde unique à la fois fluorescente et électroactive pour suivre l’exocytose par la méthodologie couplée TIRFM/ampérométrie. Ainsi, un analogue de neurotransmetteurs monoamine primaire, la 4-(2-amino-éthyl)-6-chloro-7-hydroxy-2H-1-benzopyran-2-one (nommé 1 dans ce travail), a été synthétisé.1 présente une fluorescence forte, stable et pH-dépendante. Lorsque cette entité est excitée à 405 nm, son intensité de fluorescence est presque doublée de pH 5 (valeur intra-vésiculaire) à 7 (valeur milieu extracellulaire). De plus, des études en voltammétrie ont pu mettre en évidence que 1 est oxydable sur électrode de carbone vitreux, microélectrode à fibre de carbone et ITO (oxyde d’indium dopé à l’étain), montrant ainsi une bonne électroactivité. La pénétration cellulaire dans les vésicules de cellules BON N13 a également été démontrée, prouvant la spécificité de l’interaction entre 1 et ces vésicules équipées d’un transporteur de monoamines primaires (VMAT). L’utilisation de 1 comme sonde unique optique et électrochimique pour le suivi de l'exocytose a ensuite été validée séparément dans des cellules BON N13 par TIRFM et ampérométrie. L’enregistrement simultané par fluorescence et électrochimie en utilisant 1 comme sonde double a ensuite été réalisé dans un microdispositif constitué d’électrodes ITO conductrice et transparente. Nos résultats basés sur la sonde unique 1 montrent qu’elle semble plus adaptée que toutes les stratégies antérieures impliquant deux sondes indépendantes. Les résolutions spatiale et temporelle de cette méthode combinée ont permis d'analyser des sécrétions d’exocytose de cellules marquées par 1. Une analyse ultérieure de ces signaux couplés optique et électrochimique sera à même d’étudier la corrélation entre le comportement du pore de fusion (dynamique d'ouverture/de fermeture, stabilité..) détecté par ampérométrie et le mouvement d'une vésicule en trois dimensions (ancrage, amarrage, fusion puis retrait dans le cytoplasme) détecté par TIRFM. / Vesicular exocytosis is a ubiquitous way for intercellular communications. TIRFM (total internal reflection fluorescence microscopy) and amperometry are nowadays the two most frequently used analytical methods with complementary features for its investigation. The combination of these two analytical techniques to track exocytotic secretions was firstly achieved by our group in 2011 and this new technique was demonstrated to show both high temporal and spatial resolutions by simultaneously recording the fluorescent and amperometric signals. The major disadvantage of this former work was the independent loading of optical and electrochemical probes to the secretory vesicles, which resulted in 'sightless' amperometric or optical signals as well as low coupling efficiency. Therefore, in this thesis, we attempted to develop a unique probe with dual fluorescent/electrochemical characteristics to track exocytotic process by TIRFM/amperometry coupling technique. This is why an analog of biogenic monoamine neurotransmitters, 4-(2-aminoethyl)-6-chloro-7-hydroxy-2H-1-benzopyran-2-one hydrochloride (named as 1 in this work) was synthesized. 1 exhibited bright, stable, pH-dependent fluorescence. When excited at 405 nm, its fluorescence intensity was almost doubled with the increase of pH values from 5 (similar to that in the vesicular lumen) to 7 (similar to the extracellular medium). Furthermore, in voltammetry, 1 was demonstrated to be easily electrooxidized on GCE (glassy carbon electrode), CFE (carbon fiber electrode) and ITO (indium tin oxide) electrodes surfaces, showing good electroactivity. 1 was also shown to penetrate easily into the vesicles of BON N13 cells within 1 hour incubation, testifying its specific affinity with these VMAT-equipped (vesicular monoamine transporter) vesicles. The applications of 1 as optical and electrochemical probes for exocytosis monitoring were then separately validated in BON N13 cells by TIRFM and amperometry measurements, respectively. Simultaneous recording of fluorescent and amperometric information by using 1 dual probes loaded cells was subsequently acquired in a microfabricated device constituted by conductive and transparent ITO electrodes. Our results based on the unique probe 1 for electrochemical and fluorescent detection of exocytotic release seemed more adapted than all the previous works involving independent probes. The high spatial and temporal resolutions of this combined method also allowed analyzing consecutive exocytotic secretions as well as overlapped events in 1-stained cells. Further analysis of these two signals with complementary information will shed more light on the correlation of the fusion pore behavior (opening/closure dynamics, stability…) measured by amperometry and the motion of a secretory vesicle in three dimensions (tethering, docking, fusion and retrieval) detected by TIRFM.

Philosophical and scientific investigations into the nature of belief and knowledge are ancient, extending back to the beginnings of rational thought. It is not until the last few decades that we have been able to peer into and examine the organ of belief, the brain. Neuroanatomical perspectives have begun to address the long-standing questions of epistemology by identifying specific neural regions that are critical for the storage and evaluations of beliefs. Here, a novel neuroanatomical model of belief and doubt is presented, where post-rolandic association cortices are critical for the storage of beliefs and the prefrontal cortex is necessary for the doubt and evaluation process. It is proposed that the singular function of the prefrontal cortex is "false tagging" (the neuroanatomical essence of doubt) to mental representations in post-rolandic cortices. Individuals that have dysfunction to the prefrontal cortex, such as patients with explicit damage to the prefrontal cortex, from tumor resections or cerebral vascular events, should show a "doubt deficit", accompanied by a general increased belief to information. Evidence is presented indicating that deficiencies in the "false tagging" function may explain a wide assortment of abnormalities in neurological and psychiatric patients. Several experiments in various populations (neurological, developmental, and psychiatric) were conducted to examine the role of specific brain regions in the believing and doubting process. First, two studies gave participants explicitly-labeled false beliefs and measured the ability of the participants to falsify these beliefs. It was predicted that participants with dysfunction to the prefrontal cortex would be poor at falsifying novel beliefs. Results confirmed the predictions. Second, participants were given pairs of statements that represented opposite opinions on some issue and responded by agreeing or disagreeing with each statement. Participants with dysfunction to the prefrontal cortex, who, theoretically, have a "doubt deficit," should show compartmentalized minds, where cognitions are easily believed but rarely doubted against other extant mental information. Results suggested that participants with prefrontal cortex dysfunction were more likely to agree to opposing statements. Third, individuals with dysfunction to the prefrontal cortex should lack a dissonant state that can change attitudes, when two cognitions are in conflict. Using a free-choice paradigm, it was found that participants with prefrontal dysfunction showed either extreme attitude change after choice or no attitude change after choice, which is consistent with a "no dissonance" state. Finally, individuals with compartmentalized minds tend evince an authoritarian personality. A psychometric scale and a behavioral measure of authoritarianism were examined in the participants. Results indicated that participants with prefrontal cortex dysfunction showed increased authoritarianism on the psychometric scale, but decreased authoritarian behavior, reflecting a dissociation between knowledge and behavior. In conclusion, the results support the theoretical assertions that the prefrontal cortex is critical for "false tagging" or doubting cognitive representations. Data from neurological, developmental, and psychiatric populations are broadly consistent with the theory and offer strong external validity.

While an estimated one-third of the United States population has a tattoo, tattoos are still seen as a sign of deviance. The appearance of the first tattoos in the United States were relegated to the bodies of the lower classes and outcasts of society. Over the past few decades tattoos have migrated on to the celebrity skin of today's pop culture icons. In the past twenty years, tattoos have moved from deviant subcultures to the mainstream, and yet are still considered to be a mark of the disfavored factions of society. The dominant culture continues to regard the bearers of tattoos as social deviants, while at the same time appropriating tattoos for use as fashion statements, beauty enhancements, and mechanisms for continued oppression. While tattoos make their way from the prison cell to the pop culture runway, how are they perceived by law enforcement? Are tattoos still seen as markers of deviance or has law enforcement adopted the mainstream culture's perception and view tattoos as self-expressive artwork? Do tattoos negatively influence law enforcement's judgment where individual discretion is exercised? The purpose of this study was to examine the arrest patterns of arrestees with visible tattoos using a critical theory perspective to determine if tattoos and arrest seriousness are related. This study also examines tattoo placement and type in affecting the severity of arrest charges. The data used in this study is a random sample of 2011 Pinellas County Florida arrestees (N=3,733). Numerous logistic regression models were utilized in this analysis and resulted in no consistently significant association between tattoos (visibility, placement, or type) and severity of offense charges. This provides evidence that the use of tattoos as a marker for deviance does not appear to influence police behavior any differently than other characteristics such as race.

While confessions are a powerful form of evidence, innocent people sometimes confess to crimes they did not commit. Many researchers have studied false confessions through laboratory experiments with university students or by focusing on proven cases of false confession. These approaches have led many researchers to form a conceptual framework that law enforcement interrogative methods are a key cause of false confessions. A gap exists in the literature as few researchers have queried law enforcement about false confessions or consulted with officers who specialize in interrogation. For this study, a qualitative case study approach was used to explore the experiences of 13 federal law enforcement polygraph examiners who specialize in interrogation. Telephone interviews were conducted regarding their approach to criminal interrogation, their experiences with false confessions, and the circumstances when false confessions were elicited. NVivo software was used to organize the data. Common themes in interview responses were identified and reduced to a simplified format that could be understood in the context of the research questions. The themes identified that participants conduct themselves professionally, they treat criminal subjects respectfully, they avoid unethical interrogative practices, and false confessions result from individual subject characteristics and police misconduct. No participants reported eliciting a false confession. These findings suggest that false confession researchers may have a biased view of how law enforcement officers interrogate due to their overreliance on laboratory experimentation and their focus on false confession cases. This study promotes positive social change by increasing truthful confessions, decreasing false confessions, and providing a more accurate view of what occurs during real world criminal interrogations.

L'intérêt du programme de dépistage du cancer du sein par mammographie résulte de la balance entre la réduction de la mortalité et les effets indésirables du programme. L'objectif de ce travail était d'évaluer trois aspects du programme de dépistage dans le département de l'Isère. Nous avons utilisé une méthode Approximate Bayesian Computation pour estimer le surdiagnostic lié au dépistage par mammographie de cancers non évolutifs. Ils représentaient respectivement 1,5% (IC95% : 0,3% – 2,9%) et 28,0% (IC95% : 2,2% – 59,8%) des cancers invasifs et in situ diagnostiqués. Nous avons réalisé une revue de la littérature des méthodes d'estimation du surdiagnostic et montré sur des données simulées que les méthodes basées sur les taux d'incidence annuels étaient les moins sensibles aux biais. Par ailleurs, l'incidence des cancers de l'intervalle était inférieure lorsque les mammographies étaient réalisées avec deux clichés par sein au lieu d'un seul, aussi bien à 12 que à 24 mois. Enfin, l'exclusion du diagnostic de cancer après la réalisation d'une imagerie supplémentaire, d'une biopsie ou un suivi radiologique diminuait la participation lors de la prochaine invitation au dépistage organisé. En conclusion, l'intérêt du programme de dépistage du cancer du sein par mammographie n'est pas remis en cause avec 1,4 décès évité pour un cancer surdiagnostiqué (in situ ou invasif). / The public health interest of the breast cancer screening program depends on a balance between the reduction of mortality and the side effects of the program. The goal of our work was to evaluate three particular issues of the screening program in Isère, France. We used an Approximate Bayesian Computation method to estimate overdiagnosis resulting from the detection of non progressive breast cancer. Our estimates were 1.5% (95%CI : 0.3% – 2.9%) et 28.0% (95%CI : 2.2% – 59.8%) respectively for invasive and in situ cancers. We realized a review of the methods used to estimate overdiagnosis and we found on simulated data that annual incidence methods were the least influenced by biases. In addition, the incidence of 12 months or 24 months interval cancers was lower with 2-view than with 1-view mammographies. Finally, excluding breast cancer after additional imaging evaluation, clinical and radiological follow-up, or surgical biopsy motivated by abnormal findings on index mammograms results in uniformly decreased subsequent attendance in the routine screening program. To conclude, the interest of the breast cancer screening program was not modified by our results with 1.4 life saved for one overdiagnosed breast cancer (in situ and invasive).

This collection of work contains poems that are attempting to reach a sense of understanding about the past in regard to place, landscape, architecture, and memory in relation to the self--the speaker, the self-imposed I. The perception of memory, and in particular the prevalence of false memories surrounding place and person are of major concern in this collection, along with the historical and personal narratives moving out of a voice rooted in the Mojave Desert, and in particular, Las Vegas, Nevada. These poems are working through the speaker’s complicated relationship with the desert, and the erosion of place, of home. These poems are an effort to recognize what it means to learn from the desert, to learn from Las Vegas.

Many traditional approaches cease to be useful when the number of variables is large in comparison with the sample size. Penalized regression methods have proved to be an attractive approach, both theoretically and empirically, for dealing with these problems. This thesis focuses on the development of penalized regression methods for high-dimensional variable selection. The first part of this thesis deals with problems in which the covariates possess a grouping structure that can be incorporated into the analysis to select important groups as well as important members of those groups. I introduce a framework for grouped penalization that encompasses the previously proposed group lasso and group bridge methods, sheds light on the behavior of grouped penalties, and motivates the proposal of a new method, group MCP. The second part of this thesis develops fast algorithms for fitting models with complicated penalty functions such as grouped penalization methods. These algorithms combine the idea of local approximation of penalty functions with recent research into coordinate descent algorithms to produce highly efficient numerical methods for fitting models with complicated penalties. Importantly, I show these algorithms to be both stable and linear in the dimension of the feature space, allowing them to be efficiently scaled up to very large problems. In the third part of this thesis, I extend the idea of false discovery rates to penalized regression. The Karush-Kuhn-Tucker conditions describing penalized regression estimates provide testable hypotheses involving partial residuals. I use these hypotheses to connect the previously disparate elds of multiple comparisons and penalized regression, develop estimators for the false discovery rates of methods such as the lasso and elastic net, and establish theoretical results. Finally, the methods from all three sections are studied in a number of simulations and applied to real data from gene expression and genetic association studies.

Data containing large number of variables is becoming increasingly more common and sparsity inducing penalized regression methods, such the lasso, have become a popular analysis tool for these datasets due to their ability to naturally perform variable selection. However, quantifying the importance of the variables selected by these models is a difficult task. These difficulties are compounded by the tendency for the most predictive models, for example those which were chosen using procedures like cross-validation, to include substantial amounts of noise variables with no real relationship with the outcome. To address the task of performing inference on penalized regression models, this thesis proposes false discovery rate approaches for a broad class of penalized regression models. This work includes the development of an upper bound for the number of noise variables in a model, as well as local false discovery rate approaches that quantify the likelihood of each individual selection being a false discovery. These methods are applicable to a wide range of penalties, such as the lasso, elastic net, SCAD, and MCP; a wide range of models, including linear regression, generalized linear models, and Cox proportional hazards models; and are also extended to the group regression setting under the group lasso penalty. In addition to studying these methods using numerous simulation studies, the practical utility of these methods is demonstrated using real data from several high-dimensional genome wide association studies.