There are several generally accepted rules for detecting unidimensionality, but none are well tested. This simulation study investigated well-known methods, including but not limited to, the Kaiser (k>1) Criterion, Percentage of Measure Validity (greater than 50%, 40%, or 20%), Ratio of Eigenvalues, and Kelley method, and compares these methods to each other and a new method proposed by the author (McGill method) for assessing unidimensionality. After applying principal component analysis (PCA) to the residuals of a Latent Trait Test Theory (LTTT) model, this study was able to address three purposes: determining the Type I error rates associated with various criterion values, for assessing unidimensionality; determining the Type II error rates and statistical power associated with various rules of thumb when assessing dimensionality; and, finally, determining whether more suitable criterion values could be established for the methods of the study by accounting for various characteristics of the measurement context. For those methods based on criterion values, new modified values are proposed. For those methods without criterion values for dimensionality decisions, criterion values are modeled and presented. The methods compared in this study were investigated using PCA on residuals from the Rasch model. The sample size, test length, ability distribution variability, and item distribution variability were varied and the resulting Type I and Type II error rates of each method were examined. The results imply that certain conditions can cause improper diagnoses as to the dimensionality of instruments. Adjusted methods are suggested to induce a more stable condition relative to the Type I and Type II error rates. The nearly ubiquitous Kaiser method was found to be biased towards signaling multidimensionality whether it exists or not. The modified version of the Kaiser method and the McGill method, proposed by the author were shown to be among the best at detecting unidimensionality when it was present. In short, methods that take into account changes in variables such as sample size, test length, item variability, and person variability are better than methods that use a single, static criterionvalue in decision making with respect to dimensionality. / Ph. D.

This study investigates feedback loop thinking and novices' skills in identifying feedback loops, aiming to create a valid assessment tool that distinguishes real understanding from mere awareness. The validity of the tool is assessed in two ways, 1) using it to test hypotheses about feedback learning by deploying the tool to test the effects of the terms used to name a feedback loop, and outcome desirability on loop identification accuracy determined by a signal detection analysis, and 2) using an item response analysis to determine the value of each item in distinguishing among students with a range of understanding of feedback loops. Results show little effect of terminology on sensitivity or bias, suggesting educators can use their field’s preferred terms without compromising outcomes, but a clear effect of the desirability of outcome on both sensitivity and bias, with a small increase in sensitivity and a clear bias to classify items as loops if they have a good outcome. The analysis of item sets for both positive and negative feedback loop using Item Response Theory (IRT) identified 16 items for each type of loop that provided a range of difficulties and discrimination over a moderately broad range of skill. Model fit favored a 2PL model, but global fit statistics were precluded by missing data. Still, this project delivers a reliable task for understanding feedback loop thinking, a crucial step for enabling future research on complex causal relationships. Subsequent studies should focus on assessing the unidimensionality of the item sets and elucidating the differences in how novices perceive positive versus negative feedback loops. Such efforts will further advance our understanding of feedback loop cognition. / Psychology

Statistics released in the fall 1989 showed that 238,357 (41%) of the nation's 577,710 bridges are either structurally deficient or functionally obsolete. New materials, such as fiber reinforced polymeric composites (FRP), are being suggested for use in bridge systems to solve some of the current problems. These materials are thought to be less affected by corrosive environmental conditions than conventional civil engineering materials. Therefore they may require less maintenance and provide longer life spans. More specifically, glass fiber reinforced vinyl ester matrix composites are considered possible replacements for deteriorating conventional bridge decks due to their durability, decreased weight, and relative affordability. In order to facilitate rapid acceptance of FRP structural components into the world of civil structural engineering, effective and efficient NDE techniques must be explored and documented in these situations. This thesis will discuss the use of Infrared Thermography (IRT) as a means of detecting debonds and voids caused by conditions encountered both in fabrication and in the field. As forced convective hot air is applied within the bridge deck, debonds between bridge deck components near the riding surface appear cold while imperfections near the bottom of the deck give rise to concentrations of heat. These variations in thermal propagation patterns are observed by the infrared camera and indicate possible structural deficiencies. Results of experimentation and thermal analyses from laboratory studies of a model bridge deck and some from in situ full-scale investigations are presented. / Master of Science

Previous research has been limited regarding the measurement of computational thinking, particularly as a learning progression in K-12. This study proposes to apply a multidimensional item response theory (IRT) model to a newly developed measure of computational thinking utilizing both selected response and open-ended polytomous items to establish the factorial structure of the construct, apply the recently introduced composite and structured constructs models, and to investigate the measurement invariance of the assessment between males and females using the means and covariance structures (MACS) approach.

The recent shift to a culture of competence has permeated several areas of professional psychology, including competency identification, competency-based education training, and competency assessment. A competency framework has also been applied to various programs and specialty areas within psychology, such as clinical, counseling, clinical health, school, cultural diversity, neuro-, gero-, child, and pediatric psychology. Despite the spread of competency focus throughout psychology, few standardized measures of competency assessment have been developed. To the authors' knowledge, only four published studies on measures of competency assessment in psychology currently exist. While these measures demonstrate significant steps in progressing the assessment of confidence, three of these measures were designed for use with individual programs, two of these international (i.e., UK and Taiwan). The current study applied the seminal Competency Benchmarks, via a recently adapted benchmarks form (i.e., Practicum Evaluation form; PEF), to practicum students at the University of North Texas. In addition to traditional supervisor ratings, the present study also involved self-, peer supervisor, and peer supervisee ratings to provide 360-degree evaluations. Item-response theory (IRT) was used to evaluate the psychometric properties of the PEF and inform potential revisions of this form. Supervisor ratings of competency were found to fit the Rasch model specified, lending support to use of the benchmarks framework as assessed by this form. Self- and peer-ratings were significantly correlated with supervisor ratings, indicating that there may be some utility to 360-degree evaluations. Finally, as predicted, foundational competencies were rated as significantly higher than functional competencies, and competencies improved significantly with training. Results of the current study provide clarity about the utility of the PEF and inform our understanding of practicum-level competencies.

L’étude de la qualité de vie relative à la santé est un objectif prioritaire des essais cliniques en cancérologie pour évaluer l’efficacité d’une prise en charge ; elle est mesurée par le biais d’auto-questionnaire. Dans ce travail, nous proposons différentes modélisations statistiques pour l’analyse longitudinale de ce critère, ainsi que leur application sur des données issues de plusieurs essais cliniques. Une première partie présente les modèles issus de la théorie de réponse à l’item (IRT) pour réaliser une analyse longitudinale directement sur les données brutes (multi-réponses ordinales) et ce par dimension. Une fois replacés dans le contexte des modèles linéaires généralisés mixtes, une sélection conceptuelle de modèles IRT a conclu que le Graded response model semble le mieux adapté. Dans une seconde partie, nous proposons un modèle à équation structurelle permettant de prendre en compte conjointement l’aspect multidimensionnel et longitudinal de la qualité de vie. À l’aide de facteurs reflétés par des ensembles de variables observées, il permet de lier à chaque temps de mesure toutes les observations issues du questionnaire, tout en considérant également des variables explicatives. L’analyse longitudinale est réalisée sur le statut global de santé et les facteurs réduisant ainsi le nombre de tests. Enfin, une approche par mélanges de modèles mixtes est proposée pour obtenir des classes latentes à partir de trajectoires de qualité de vie. Cette approche a permis de caractériser des sous-populations homogènes et d’associer différente évolution de la qualité de vie suivant des profils particuliers de patients. / The health-related quality of life is a major objective in oncology clinical trials to improve patients’ care and better evaluate the impact of the treatments on their everyday life. Auto-questionnaires are usually used to measure this endpoint. In this work, different statistical models for the longitudinal analysis of health-related quality of life in oncology are proposed and applied to clinical trial data. First, we present different models derived from the item response theory (IRT) to achieve a longitudinal analysis directly on raw data (multi-response outcomes) for each dimension. Within the generalized linear mixed model background, a conceptual selection of the IRT models concluded that the graded response model seems to be the most suitable. Then, we propose a structural equation model which allows taking into account the multidimensional nature of data at each time and the longitudinal aspect induced by the repeated measurements. At each measurement time, the model allows to link all the observed variables issued from the questionnaire given explanatory variables. Two factors are estimated, each summarizing a set of observed variables. The longitudinal analysis is performed on the global health status and on the factors, thus reducing the number of tests. Finally, an approach based on a mixture of mixed models is used to obtain latent classes from quality of life trajectories. The approach has resulted in the identification of homogeneous subpopulations and their latent trajectory according to specific patient profiles.

Le présent travail de thèse fait suite à une première étude où une stratégie d’identification des paramètres et formes des lois de zones cohésives a été élaborée pour des matériaux homogènes. L’extension au cas de matériaux présentant des hétérogénéités nécessite d’accéder localement aux champs de contraintes.Ainsi, l’objectif principal de cette étude est de mettre au point une méthode de caractérisation locale des propriétés mécaniques et des contraintes. Cette méthode est basée sur l’erreur en relation de comportement combinée à l’exploitation de la richesse des mesures de champs cinématiques planes et plus particulièrement des champs de déformations, obtenus par dérivation numérique des champs de déplacements. Cette mesure cinématique est réalisée par une technique de corrélation d’images numériques enrichie.La méthode d’identification est basée sur la minimisation itérative d’une norme énergétique faisant intervenir le tenseur élastoplastique sécant. Différentes simulations numériques ont illustré la capacité de la procédure à identifier localement des champs de propriétés hétérogènes et sa robustesse et sa stabilité vis-à-vis du bruit de mesure, du choix du jeu de paramètres d’initialisation de l’algorithme et de la finesse du maillage.Pour finir, des essais plans avec différentes géométries d’éprouvettes ont été effectués et un essai a été mis au point pour obtenir de manière maîtrisée un état initial très hétérogène. Les résultats d’identification élastoplastique multilinéaire ont montré la capacité de la méthode à identifier les lois de comportements locales sur ce matériau hétérogène. / The present work follows a first approach where a strategy for identifying the shape and the parameters of cohesive-zone laws has been developed for homogeneous materials. The extension of this method to heterogeneous material requires the knowledge of the local stress state.The study aims at developing a local characterization method for mechanical properties and stresses. This method is based on the constitutive equation gap principles and relies on the knowledge of mechanical kinematic fields and particularly of the strain fields. These fields are obtained by the numerical differentiation of displacement fields measured by digital image correlation.This identification method is based on the iterative minimization of an energy norm involving the secant elastoplastic tensor. Various numerical simulations were used to illustrate the performance of the procedure for locally identifying heterogeneous property fields, and to characterize its robustness and its stability with respect to noise to the values of the algorithm initialization parameter and to the mesh refinement.Finally, various experimental tests with different specimen geometries were performed and a test has been developed to obtain a controlled heterogeneous initial state. The multilinear elastoplastic identification results showed the ability of the method to identify the local behavior properties on heterogeneous materials.

隨著網路的發達,各式各樣的資訊和商品也在網路上充斥著,使用者尋找資訊或是上網購物時,有的網站有推薦系統(recommender system)能提供使用者相關資訊或商品。若推薦系統能夠讓消費者所搜尋的相關資訊或商品能夠符合他們的習性時,便能讓消費者增加對系統的信賴程度,因此系統是否能準確預測出使用者的偏好就成為一個重要的課題。本研究使用兩筆資料,並以相關研究的三篇文獻進行分析和比較。這三篇文獻分別為IRT模型法(IRT model-based method)、相關係數法(correlation-coefficient method)、以及矩陣分解法(matrix factorization)。 在經過一連串的實證分析後,歸納出以下結論: 1. 模型法在預測方面雖然精確度不如其他兩種方法來的好,但是模型有解釋變數之間的關係以及預測機率的圖表展示,因此這個方法仍有存在的價值。 2. 相關係數法容易因為評分稀疏性的問題而無法預測,建議可以搭配內容式推薦系統的運作方式協助推薦。 3. 矩陣分解法在預測上雖然比IRT模型法還好，但分量的數字只是一個最佳化的結果，實際上無法解釋這些分量和數字的意義。 / With the growth of the internet, websites are full of a variety of information and products. When users find the information or surf the internet to shopping, some websites provide users recommender system to find with which related. Hence, whether the recommender system can predict the users' preference is an important topic. This study used two data,which are "Mondo" and "MovieLens", and we used three related references to analyze and compare them. The three references are following: IRT model-based method, Correlation-coefficient method, and Matrix factorization. After the data analysis, we get the following conclusions: 1. IRT model-based method is worse then other methods in predicting, but it can explain the relationship of variables and display the graph of predicting probabilities. Hence this method still has it's value. 2. Correlation-coefficient method is hard to predict because of sparsity. We can connect it with content filtering approach. 3. Although matrix factorization is better then IRT model-based method in predicting, the vectors is a result of optimization. It may be hard to explain the meaning of the vectors.

Syftet med studien var att undersöka hur Parken Zoo använde sig av Facebook för att kommunicera med sina intressenter under en kris som drabbade organisationen hösten 2012. Med hjälp av en kvalitativ textanalys studerade vi de Facebookinlägg som Parken Zoo hade publicerat under krisen. Därefter genomförde vi en kvalitativ intervju med marknadschefen på Parken Zoo, som var ansvarig för kommunikationen på Facebook under krisen. Med hjälp av ett analysverktyg baserat på situational crisis communication theory och image restoration theory kunde vi avgöra vilka försvarsstrategier Parken Zoo hade använt i Facebookinläggen. Resultatet visade att Parken Zoo omedvetet hade använt många olika försvarsstrategier, men att tre tydliga budskap förmedlades i Facebookinläggen. Resultatet visade också att Parken Zoos syfte var att använda Facebook som ett dialogverktyg, men att de i själva verket använde Facebook för att sprida enkelriktad kommunikation. En slutsats vi kunde dra från studen var att det är viktigt för organisationer att vårda sina kundrelationer genom att bemöta intressenternas frågor och kommentarer på sin Facebooksida under en kris. En annan slutsats var att det är viktigt för organisationer att ha en god krisberedskap och att inkludera Facebook i den. / The purpose of this study was to examine how Parken Zoo used Facebook to communicatate with their stakeholders during a crisis 2012. Using a qualitative textual analysis, we examined Parken Zoo’s Face- book posts during the crisis. Next, we conducted a qualitative interview with the marketing manager at Parken Zoo, who was responsible for the organization’s communication on Facebook during the crisis. Using an analysis tool based on situational crisis communication theory and image restoration theory, we assessed which crisis response strategies Parken Zoo had used in their Facebook posts. The results showed that Parken Zoo subconsciously had used many different crisis response strategies, but that three clear messages were conveyed in the Facebook posts. The results also showed that Parken Zoo’s aim was to use Facebook as a dialogue tool, but in fact they used it for one-way communication. One conclusion from this study was that it is important for organizations to care for their relationship with the stakeholders by responding to their questions and comments on their Facebook page during a crisis. Another conclusion was that it is important for organizations to develop a proactive crisis management portfolio and to include Facebook in it.

Ce travail s’inscrit dans un projet de recherche interdisciplinaire (Zone Atelier Bassin du Rhône – Agence de l’Eau Rhône Méditerranée Corse) portant sur la rivière Cèze, affluent du Rhône.Le terrain d’expérimentation se situe dans les formations karstiques du bassin de la Cèze (Gard, France). Cette zone touristique est exposée à une croissance démographique et de l’activité agricole, engendrant une augmentation de la demande en eau. La thèse se concentre sur la restitution des eaux karstiques à la rivière en période estivale pour en comprendre le fonctionnement de l’hydrosystème en période de basses eaux, quand la ressource est vulnérable.Ce travail a permis d’élaborer une méthodologie, pour analyser et quantifier les échanges entre la rivière et l’aquifère karstique, fondée sur : la géologie, l’hydrologie, la géochimie, la biologie, la radioactivité en radon, l’analyse d’images infrarouges thermiques et la modélisation. Les résultats obtenus avec ces approches sont confrontés pour interpréter les interactions karst/rivière d’un point de vue qualitatif et/ou quantitatif (localisation, périodicité, débits). La confrontation de ces résultats met en avant l’intérêt d’une méthodologie interdisciplinaire pour interpréter et quantifier les échanges karst/rivière. L’application de la méthode montre qu’en juin 2015, la Cèze est alimentée à 50 % par des eaux karstiques.L’analyse multi-métrique du système karstique a permis d’acquérir de nouvelles connaissances sur son fonctionnement nécessaires pour paramétrer le modèle par réseaux de neurones qui constitue la dernière étape de ce travail. / This work is part of an interdisciplinary research project (Rhone Basin Workshop Zone – the Rhone-Mediterranean and Corsica Water Agency) on the river Cèze, a tributary of the Rhône.The experimental field is located in the karstic formations of the Cèze basin (Gard, France). This tourist area is exposed to population growth and agricultural activity, causing an increase in water demand. The thesis focuses on the karstic water restitution to the river during summer, in order to understand the functioning of the hydrosystem in periods of low water levels, when the resource is vulnerable.This work led to the development of a methodology to analyze and quantify the exchanges between karstic aquifers and rivers. This methodology is based on geology, hydrology, geochemistry, biology, radon radioactivity, infrared thermal imaging analysis and modeling. The results obtained with these approaches are compared in order to understand the karst/river interactions from a qualitative and/or quantitative point of view (localization, frequency, flow rates). The comparison of these results highlights the advantages of an interdisciplinary methodology for understanding and quantifying the karst/river exchanges. The application of this method shows that in June 2015, 50 % of the river Cèze was fed by karstic waters.The multi-metric analysis of the karstic system has led to new knowledge about its functioning. This knowledge is necessary to set the model’s parameters using neural networks, which is the last stage of this work.