Observational studies are an indispensable complement to randomized clinical trials (RCT) for comparison of treatment effectiveness. Often RCTs cannot be carried out due to the costs of the trial, ethical questions and rarity of the outcome. When noncompliance and missing data are prevalent, RCTs become more like observational studies. The main problem is to adjust for the selection bias in the observational study. One increasingly used method is propensity-score matching. Compared to traditional multi-covariate matching methods, matching on the propensity score alleviates the curse of dimensionality. It allows investigators to balance multiple covariate distributions between treatment groups by matching on a single score. This thesis focuses on the large sample properties of the matching estimators of the treatment effect. The first part of this thesis deals with problems of the analytic supports of the logit propensity score and various matching methods. The second part of this thesis focuses on the matching estimators of additive and multiplicative treatment effects. We derive the asymptotic order of the biases and asymptotic distributions of the matching estimators. We also derive the large sample variance estimators for the treatment effect estimators. The methods and theoretical results are applied and checked in a series of simulation studies. The third part of this thesis is devoted to a comparison between propensity-score matching and multiple linear regression using simulation.

There are no previous studies validating the accuracy and repeatability of ClearMatch photographic shade analysis system. The purpose of this study was to compare the shade matches performed by a photographic shade analysis system (ClearMatch) with conventional visual shade matching method under simulated clinical conditions. Three shade matching methods were used to match twelve shade tabs under simulated clinical conditions using a Vita Classical shade guide: conventional visual shade matching using 3 human raters (VM – Visual method), photographic shade analysis system (CM - ClearMatch) using two different ways of normalizing the image (CM-A2 and CM-ref respectively). Shade matching for all methods was completed at two separate sessions. The Kappa statistic was used to determine the intra-rater and inter-rater agreement. CIELAB values of the shade results were used to produce scatter plots as well as to calculate the color difference (delta E) between VM and CM groups. There was no inter-rater agreement between VM and CM-A2 (k=0.000 and k=0.015 for the first and second sessions respectively) while VM and CM-ref showed weak agreement (k=0.244 and k=0.091 respectively). Intra-rater agreement was strong in all groups VM, CM-A2 and CM-ref (0.705, 0.803 and 0.681 respectively). CM-A2 had 2 (8.3%) shade matches with a delta E of less than 2.6 (clinically imperceptible), while CM-ref had 12 (50.0%) imperceptible matches. CM-A2 had an additional 16 (66.6%) shade matches with delta E of less than 5.5 (clinically acceptable), while CM-ref had 23 (95.8%) additional acceptable matches. There was poor agreement in exact shade matches between conventional visual shade matching method and the photographic shade analysis system. The repeatability of the photographic shade analysis system was shown to be comparable to conventional visual shade matching. Using conventional shade matching as the gold standard, the capability of this photographic shade analysis system to accurately shade match has not been achieved.

In Experiment 1, four subjects were trained to match two visual samples (A) and their respective nonidentical visual comparisons (B); i.e., A-B matching. During nonreinforced test trials, all subjects demonstrated stimulus equivalences within the context of sample-comparison reversibility (B-A matching): When B stimuli were used as samples, appropriate responding to A comparisons occurred. A-B and B-A matching persisted given novel stimuli as alternate comparisons. However, the novel comparisons were consistently selected in the presence of nonmatching stimuli: i.e., during trials comprised of a novel comparison, an A or B sample from one stimulus class, and an "incorrect" comparison from the other, B or A stimuli respectively. In Experiment 2, three groups of subjects were trained under three different mediated transfer paradigms (e.g., A-B, C-B matching). Tests for reversibility (e.g., B0A, B0C matching) and mediated transfer (e.g., A-C, C-A matching)evinced stimulus equivalences for 11 of 12 subjects. The 11 subjects also matched the mediated equivalences given novel comparisons; whereas, they selected the novel comparisons when combined with nonmatching stimuli. Overall, the demonstrated stimulus equivalences favor a concept learning interpretation of non-identity matching-to-sample. Additionally, the trained and mediated matching relations were comprised of complementary sets of S+ and S- rules: Any stimulus of a given class used as a sample designated both the "correct" and "incorrect" comparisons.

The purpose of this study was to develop behavior by the pigeon illustrative of S+ and S- rules and to examine if behavior consistent with a concept rule interpretation developed. In order to examine this possibility six groups of pigeons (N=31) were provided different histories of autoshaping. Histories involved the identity of the color of the lighted center key and one side key of three horizontally mounted pigeon keys. Center key onset was followed three seconds later by onset of either outer key. Outer key onset was followed six seconds later by food presentation (explicitly paired) or 45 seconds later, during the inter-trial interval (explicitly unpaired). The foregoing reinforcement conditions were factored into two stimulus conditions, one where the center and side keys were lit by the same hue (identity) and one where the center and side key were lit by different hues (nonidentity). Two groups received identity stimulus sequences with explicitly unpaired food presentation. Two groups were exposed to nonidentity stimulus sequences with explicitly unpaired food presentation. One group received sessions combining exposure to both the explicitly paired identity and explicitly unpaired nonidentity trials. One group received no pretraining. Following pretraining, all birds were placed in a simultaneous matching-to-sample task utilizing the same hues used during pretraining. After reaching criterion on matching-to-sample, on a random ratio 3 schedule of reinforcement, birds were exposed to transfer tests, with a novel hue, to assess for S+ rules, S- rules, and a concept rule. The results indicated that there were no significant differences among groups in terms of their acquisition of matching-to-sample or in terms of their performance on transfer tests. All birds' responding during transfer conditions provided evidence of S+ rules, but neither demonstration of S- rules nor concept rule performance was evidenced. During autoshaping, birds in the identity, explicitly paired groups responded primarily to the center key, suggesting that the stimulus on the outer key was not a salient stimulus for the identity discrimination. For birds in the nonidentity, explicitly unpaired groups neither the center nor outer key controlled responding. For the group combining identity, explicitly paired and nonidentity, explicitly unpaired trials, the birds failed to form a discrimination between types of trials.

This thesis is an examination of a firearm identification system that creates a three-dimensional image of a bullet in order to record the depth and length of striations occurring along the bullet's surface. Ballistics evidence is an area of forensics in great need of further development. The advent of more sophisticated firearms such as semi-automatic and automatic weapons has increased the need for a matching system that connects bullets found at crime scenes with suspect guns. In the past, control bullets matching ones found at the crime scene have been test fired and then examined by a comparison microscope for similarities with the evidence bullet. The purpose of this thesis is to examine data collected by an emerging system that uses three-dimensional technology by way of a laser and convex mirrors to create a digitized representation of the lands and grooves of a bullet. This representation is a measure of the depth of striations or markings created on the bullet's surface during the firing event. The objective of this thesis is to statistically examine the data collected by this system, which consists of bullets produced by eight different manufacturers. The data for this thesis comes from a pilot study conducted by the creators of a three-dimensional system called SCICLOPS. Variables examined include the maximum and minimum number of striations recorded, the relative position of the bullet (as determined by the six lands and grooves measured by the system), and the manufacturer type. It is hypothesized that there will be differences in the number of striations measured across manufacturer types. Results indicate that manufacturer type may play an important role in how bullets "take" striations or markings during the firing event. Implications for the SCICLOPS system and future research are discussed.

There have been many theories about cultural differences found between groups on intelligence test scores. The main debate has been between those in favor of a genetic explanation versus those in favor of a more environmental one. When considering environmental influences, one explanation has been that there could be differential effects due to the assessor. Although there have been several studies that have considered this possibility, the results are inconclusive. The current study attempted to tease apart the assessor effects by focusing on biases in the assessor alone and by eliminating effects from the test taker. The study is an experimental design where participants were randomly assigned a WISC-III protocol of members of different ethnic groups. It was hypothesized that different groups may score these IQ tests differentially depending on the race/ethnicity of the person who was assessed. Results showed that when given identical protocols, participants scored African American protocols lower than Caucasian American protocols in both high and average IQ conditions. Clinical implications of these results are discussed.

This work aims to enhance the Ensemble Kalman Filter performance by transforming the non-Gaussian state variables into Gaussian variables to be a step closer to optimality. This is done by using univariate and multivariate Box-Cox transformation. Some History matching methods such as Kalman filter, particle filter and the ensemble Kalman filter are reviewed and applied to a test case in the reservoir application. The key idea is to apply the transformation before the update step and then transform back after applying the Kalman correction. In general, the results of the multivariate method was promising, despite the fact it over-estimated some variables.

A self-control procedure that involved a later onset of a stimulus signalling a small reinforcer within the waiting time for a larger reinforcer was investigated to determine a point of shifting preference and a discounting function as the delay varied. The results from Experiment 1 to Experiment 3 showed exclusive impulsive choices regardless of the delay. In order to examine if the results were due to the procedure and the parameters, or the species used, Experiment 4 attempted to obtain shifts in preference using simultaneous onset of stimuli with the same species. The results demonstrated no changes in preference but an increase in proportion of self-control choices was shown. Due to the limited information from the replicated studies, the accounts for the results could not be concluded. The explanations derived from choice models seemed most plausible, but limitations of the choice models were discussed.

A consecutive number of studies on the adoption trend of logistics technology since 1988 revealed that logistics organizations are not in the frontier when it comes to adopting new technology and this delayed adoption creates an information gap. In the advent of supply chain management and the strategic position of logistics, the need for accurate and timely information to accompany the logistics executives became more important than ever before. Given the integrative nature of logistics technology, failure to implement the technology successfully could result in writing off major investments in developing and implementing the technology or even in abandoning the strategic initiatives underpinned by these innovations. Consequently, the need to employ effective strategies and models to cope with these uncertainties is rather crucial. This thesis addresses the aspect of uncertainty in implementation success by process and factor research models. Process research approach focuses on the sequence of events in the technology transfer process that occurs over time. It explains the story that explains the degree of association between these sequences and implementation success. Through content analysis, this research gathers, extracts, and categorizes process data of actual stories of logistics technology adoption and implementations in organizations that are published in literature. The extracted event sequences are then analyzed using optimal matching from natural science and grouped using cluster analysis. Four patterns were revealed that organizations follow to transfer logistics technology namely, formal minimalist, mutual adaptation, development concerned, and organizational roles dispenser. Factors that contribute to successful implementation in each pattern were defined as the crucial and necessary events that characterized and differentiated each pattern from others. The factor approach identifies the potential predictors of successful technology implementation and tests empirical association between predictors and outcomes. This research develops a logistics technology success model. In developing the model, various streams of research were investigated including logistics, information systems, and organizational psychology. The model is tested using a questionnaire survey study. The data were collected from Australian companies which have recently adopted and implemented logistics technology. The results of a partial least squares structured equation modeling provide strong support for the model constructs and valuable insights to logistics/supply chain managers. The last study reports a convergent triangulation study using multiple case study of three Australian companies which have implemented logistics technology. A within and a cross case analysis of the three cases provide cross validation for the results of the other two studies. The results provided high predictive validity for the two models. Furthermore, the case study approach was so beneficial in explaining and contextualizing the linkages of the factor-based model and in confirming the importance of the crucial events in the process-based model. The thesis concludes with a research and managerial implications chapter which is devoted for logistics/supply chain managers and researchers.

Cette thèse présente la fusion multi-capteurs par un réseau Bayésien appliqué au problème de localisation d'un véhicule sur une carte. La mise en correspondance d'une estimation sur un segment de route ou Road-matching consiste à trouver le segment sur lequel le véhicule roule et la position de ce véhicule sur ce segment. Plusieurs algorithmes utilisent la fusion des estimations données par l'odométrie et le GPS pour traiter le problème du road-matching. Cependant, une simple combinaison du GPS et de l'odométrie ne permet pas de se localiser de manière précise et sans interruption de service. La précision et la continuité de service peuvent être améliorées si on utilise des informations cartographiques qui permettent en particulier de contraindre les positions possibles aux seuls segments correspondants à des voies de circulation autorisées.\\ Dans de nombreux cas, lorsqu'un véhicule se trouve devant des situations ambiguës comme les routes parallèles, les jonctions de routes,... plusieurs auteurs cherchent à sélectionner le segment le plus probable. Cette phase est souvent une source d'erreurs. Dans cette thèse nous proposons de traiter tous les segments candidats jusqu'à la levée de l'ambiguïté. \\ Le problème de la localisation devient encore plus compliqué quand il s'agit d'une localisation multi-véhicules pour une navigation autonome des véhicules suiveurs. Pour un train de véhicules dont seul le premier est piloté par un opérateur humain et dont les véhicules suiveurs sont en mode autopilotage, une géo-localisation précise d'ordre centimétrique de chaque véhicule est plus que nécessaire pour les modules de contrôle pour le suivi de trajectoire du véhicule de tête. \\ Un train de véhicule peut être vu comme la généralisation du modèle de réseau Bayésien pour la localisation d'un véhicule sur une carte. Nous dupliquons le réseau autant de fois qu'on a de véhicule. Nous rajoutons des liens de connexions entre les véhicules afin de concevoir le train de véhicule.\\ Le filtre de Kalman et sa version étendue sont très utilisés en robotique, principalement pour traiter le problème de la non linéarité. Cependant, la linéarisation du système autour de l'estimée courante peut introduire des erreurs sur la moyenne et la covariance calculées \textit{a posteriori} et peut même dans d'autres cas faire diverger le filtre. \\ La transformation du système non linéaire d'un véhicule sous forme chaînée permet de représenter son équation cinématique sous forme linéaire. Par conséquent, cette transformation nous évite de faire appel aux méthodes d'inférence approximatives.