Universidade Federal do Cearà / Este trabalho tem o objetivo de desenvolver um estudo de caso sobre o fenÃmeno da evasÃo no ensino superior de uma universidade em particular. No Ãmbito internacional, a evasÃo no ensino superior tem sido bastante discutida. No Brasil, esta temÃtica vem sendo objeto de estudo cientÃfico desde a dÃcada de 80, e em mÃdia verificamos uma grande quantidade de estudos particularizados. A partir dos dados obtidos junto a uma InstituiÃÃo de Ensino Superior, a Universidade de Fortaleza â Unifor, situada na cidade de Fortaleza - CearÃ, esta dissertaÃÃo propÃs estudar o fato do aluno evadir ou nÃo o seu curso de graduaÃÃo. O mÃtodo utilizado para o desenvolvimento do trabalho foi uma coleta de dados proveniente do sistema acadÃmico e questionÃrio sÃcio econÃmico, no qual configurados em uma planilha matriz foi possÃvel diagnosticar o fenÃmeno da evasÃo. Este trabalho desenvolveu e aplicou uma metodologia baseada no modelo de regressÃo logÃstica com o intuito de observar a evasÃo no ensino superior e o perÃodo estudado foi 1994.2 atà 2004.1. O resultado encontrado nos revelou que a variÃvel dummy evasÃo està diretamente relacionada as variÃveis independente sexo, estado civil, escolaridade do aluno, remuneraÃÃo do aluno e os centros dos cursos de graduaÃÃo. Entretanto, as variÃveis sexo e renda do aluno foram as estatisticamente mais significativas no processo de decisÃo do aluno evadir ou nÃo. Para reduzir a evasÃo, a universidade deverà implantar um sistema que possa avaliar a satisfaÃÃo do aluno em relaÃÃo à universidade como, tambÃm, buscar identificar quais sÃo as maiores dificuldades encontradas pelos alunos a fim de reduzir a evasÃo dos mesmos. / This work aimed to produce a case study about evasion of undergraduate students in a particular university. Data was obtained from the Universidade de Fortaleza - Unifor, located at Fortaleza, state of CearÃ, Brazil. This dissertation was proposed to investigate whether the student evades or not his/her undergratuate course. To achieve its goal, the work developed a logistic econometric model by which it was observed that evasion is related directly to independent variables: sex, civil states, studentâs scholarship, studentâs earnings, areas of knowledge in this case study.

Chlorine is always used in disinfection process of water treatment plant (WTP) in Kaohsiung, and the major of disinfection by-products (DBPs) is THMs. Reaction of THMs producing were found in relation to lots of factors of water quality. Therefore we used SPSS to analysis the relation between water quality and THMs content for making prediction model of THMs formation in water distribution system in downtown Kaohsiung city. It is expected to estimate the concentration of THMs with a simple method, and to reduce THMs by controlling some items of water quality. While the variance of TOC value is low, we found that concentration of THMs is related to temperature, pH, and residual chlorine by regression method. In these items, residual chlorine had the highest correlation. While the variance of TOC value is high, the correlation between the three items and THMs is weaker than the front one. From this finding, TOC is best one fitting factor in prediction of THMs. In this study, we found that R-square value of Multivariate regression prediction is higher than the value of Univariate regression prediction probably for THMs producing is a complex reaction, so formation model was hard using only one item of water quality.

This study responds to the globally increasing rate of caesarean section, and specifically to the very high rate of elective caesarean section among Taiwanese mothers as evidence suggests that such elective caesareans pose potential health risks for mothers and babies. The purpose of this study was to develop and evaluate a multi-component instrument based on the theory of planned behavior (TPB) to better understand Taiwanese pregnant women's decisions regarding their childbirth delivery options (spontaneous vaginal delivery or elective caesarean section). The study was a four-phased mixed method design. First, the TPB guided item development and instrument drafting. Second, pretesting and instrument refinement used cognitive interviewing with a small sample of Taiwanese pregnant women. Third, the instrument was administered to 310 such women to examine psychometric properties of the component scales. Fourth, the phase 3 instrument was re-administered to 30 women to estimate item stability. Confirmatory factor analyses (CFA) were used to assess construct validity of the multi-item, multi-component measurement model with LISREL 9.1. Based on the TPB, the 52-item self-administered Childbirth Delivery Options Questionnaire (CDOQ) was developed to measure three components: intention regarding delivery options, attitudes toward delivery options, and perceptions of significant others' (partner, mother, and mother-in-law) feelings about delivery options. Respondents from phase two thought that the items on the CDOQ were easy to read and comprehend; they reported favorably on the wording and formatting. Preliminary item analysis revealed that the items referring to dangerousness of delivery options did not function as intended and were dropped because they did not differentiate between the two delivery options, leaving 36 items. Test-retest reliability indicated that responses to each item were positively correlated and those referring to spontaneous vaginal delivery were more stable than those referring to elective caesarean section. Corrected item-to-total correlations and expected change in Cronbach's alpha if item deleted revealed that four items might form a measure of general social norms associated with the Taiwanese culture. The Cronbach's alphas for the components of the CDOQ ranged from .55 to .89. The measurement model incorporating the design features of the CDOQ fitted the data well using the CFA. Because serious problems with multicollinearity and suppression were revealed, Beckstead's (2012) criterion-irrelevant-variance-omitted (CIVO) regression method was used to untangle the suppressor effects when predicting intention from the other components of the CDOQ. The results indicated that attitude and partner's feelings were significant and explained the bulk of the variance in intentions. The TPB-based instrument developed here will be of considerable use to maternal-child health researchers. The findings of this study suggest that decisions regarding delivery options may be modified by interventions geared toward pregnant women's attitudes within family- and cultural-centered prenatal programs.

The paper consists of two parts. In the first part the linear regression method is described, also employment possibilities and pupil achievement research are discussed. The second one deals with some factors and what influence they have on the eight class pupil’s mathematical achievement analysis.

Conselho Nacional de Desenvolvimento Científico e Tecnológico / Phosphorus (P) is essential for swine and must be supplied in adequate amounts in order to minimize the costs of production and to reduce environmental impact. The standardized digestibility is a good estimate of available P on feedstuffs. To determine the standardized total tract digestibility of P (STTD) it is necessary to know the endogenous losses of P (ELP). The P-free diet is a methodology largely utilized to find ELP. Gelatin (GEL) is the elected protein source to formulate a P-free diet in most trials around the world. In Brazil, however, spray dried plasma (SDP) has been utilized as a protein source in P-free diets. In spite of small content and high digestibility of P present in SDP, its use may lead to incorrect values of ELP and, as a consequence, overestimation of the standardized digestibility coefficients of P in feedstuffs for pigs. However, one of the advantages of SDP use is the small cost in relation to GEL and, for this reason, more studies on the influence of SDP on ELP and P digestibility are justified. Therefore, a trial was carried out to compare the ELP in swine fed diets containing GEL or SDP as a protein source and to determine the STTD in SDP. The study was developed at the Swine Farm of the Federal University of Santa Maria Brazil. A total of 12 castrated pigs with average initial weight of 55 kg were individually allotted to metabolic crates. The trial was developed in two 12-days periods, with 7 days of adaptation and 5 days of total fecal collection. The beginning and the end of the collection were determined according to the marker-to-marker approach, using ferric oxide as an indigestible marker. Pigs were weighed at the start and at the end of adaptation period and at the end of collection. The treatments were four semi-purified diets. One of them had a P-free diet with 30% of GEL as the only protein source. The other three treatments had 10, 20 and 30% inclusion of SDP. A 1:1 Ca to P ratio was used for all diets. Data were subjected to ANOVA and the model included the effects of period, animal and treatments. Then, the results of the three diets with increased levels of SDP were subjected to linear regression analysis. The intercept of the relation of ingested P and absorbed P represented the endogenous losses of P. The slope indicated the STTD of SDP. The ELP means obtained by P-free diet and regression method were compared with the Student t test. The ELP were 128.95 mg/kgDMI and 153 mg/kgDMI (SE = 77.0; P<0.06) using the P-free diet with GEL as the protein source and the regression method, obtained with diets containing increased levels of SDP, respectively. The apparent digestibility of P was 87.9, 94.2 and 92.9% for the treatments containing 10, 20 and 30% inclusion of P, respectively. The estimated STTD obtained with the linear regression was 97.4%. The corrected values of apparent digestibility of P, obtained with the diets containing SDP, using the basal endogenous loss estimated by the P-free diet, were 96.9, 98.8 and 95.9% for 10, 20 and 30% of SDP, respectively. Therefore, it can be concluded that SDP can replace GEL as a source of protein (amino acids) in P-free diets to estimate the endogenous losses of P. In addition, the standardized digestibility of P in SDP estimated with the P-free diet with GEL as the protein source was 97.2% and it was 97.4% obtained by the regression method, utilizing SDP as the source of protein. / O fósforo (P) é essencial aos suínos e precisa ser fornecido em quantidades adequadas para minimizar os custos de produção e reduzir o impacto ambiental. A digestibilidade padronizada é considerada uma boa estimativa do P útil presente nos alimentos destinados aos suínos. Para a determinação da digestibilidade total padronizada de fósforo (DTPP) é necessário conhecer as perdas endógenas de fósforo. Uma das metodologias mais utilizadas para encontrar as perdas endógenas de P é a da dieta isenta de fósforo. O uso da gelatina (GEL) como ingrediente proteico em dieta isenta de P tem sido adotado por pesquisadores. No Brasil, o plasma sanguíneo desidratado (PS) vem sendo utilizado como fonte proteica em dietas isentas de fósforo. Apesar do baixo conteúdo e a alta digestibilidade do P presente no PS, seu emprego pode gerar valores incorretos de perdas endógenas de P e, em decorrência, superestimar os coeficientes de digestibilidade padronizada do P nos alimentos de suínos. Contudo, uma das vantagens da utilização do PS é seu custo inferior em relação a GEL e, por essa razão, estudos mais aprofundados da influência do PS nas perdas endógenas e digestibilidade de P no PS se justificam. Dessa forma, um experimento foi realizado com o objetivo de comparar as perdas endógenas de P em suínos alimentados com dietas contendo GEL ou PS como fonte proteica e também determinar a DTPP no plasma sanguíneo desidratado. O experimento foi realizado no Setor de Suínos da Universidade Federal de Santa Maria RS. Utilizaram-se 12 suínos machos castrados com peso médio inicial de 55kg, alojados individualmente em gaiolas de metabolismo. O experimento foi constituído por dois períodos de 12 dias, sendo que cada período compreendeu sete dias destinados à adaptação e cinco para a coleta total de fezes. O início e o final do período de coleta foram determinados pelo aparecimento de fezes marcadas devido a adição de óxido férrico nas dietas. Os animais foram pesados no início e final da adaptação e ao final do período de coleta. Foram distribuídos em quatro tratamentos constituídos por dietas semipurificadas. Um dos tratamentos foi constituído por dieta isenta de P e teve 30% de incorporação de GEL, sendo esta a única fonte proteica da dieta. Os outros três tratamentos utilizaram uma dieta com teores de 10, 20 e 30% de inclusão de plasma sanguíneo desidratado. A relação entre cálcio e P total das dietas foi de 1:1. Os dados foram submetidos a análise de variância utilizado no modelo o efeito do período, animal e tratamentos. Posteriormente, os resultados das três dietas com níveis crescentes de inclusão de PS foram submetidos à análise de regressão linear, sendo que o intercepto da relação entre fósforo ingerido (PING) e fósforo absorvido (PABS) representou as perdas endógenas de P e o coeficiente angular da reta indicou a DTPP do plasma sanguíneo desidratado. As estimativas de perdas endógenas obtidas pela dieta isenta de P e pelo método da regressão foram comparados pelo teste t de Student. As perdas endógenas de P da dieta isenta de P foram de 128,95mg/kgMSING e 153 mg/kgMSING (EP=77,0; P<0,06) com a dieta isenta de P, tendo a GEL como fonte de proteína, e pelo método de regressão, estimada com dietas contendo níveis crescentes de PS, respectivamente. A digestibilidade aparente do P foi de 87,9, 94,2 e 92,9% para os tratamentos com 10, 20 e 30% de inclusão de fósforo, respectivamente. Já a estimativa da digestibilidade padronizada de P gerada pela técnica da regressão linear simples foi de 97,4%. A correção dos valores de digestibilidade aparente do P, obtidos com as dietas contendo PS, usando a perda endógena basal estimada pelo método da dieta isenta de P, resultou em digestibilidade total padronizada de 96,9, 98,7 e 95,9% para o P para os níveis de inclusão de PS de 10, 20 e 30%, respectivamente. Com isso, pode-se concluir que o plasma sanguíneo desidratado pode substituir a gelatina como fonte de proteína em dietas isentas de fósforo elaboradas para determinar perdas endógenas de fósforo e a digestibilidade padronizada do fósforo no plasma sanguíneo desidratado foi de 97,2%, estimada pela dieta isenta de fósforo, tendo a gelatina como fonte de proteína e 97,4% pelo método da regressão, utilizando plasma sanguíneo desidratado como fonte de proteína.

Doctor of Philosophy / Department of Civil Engineering / Sunanda Dissanayake / Lane-departure crashes are the most predominant crash type in Kansas which causes very high number of motor vehicle fatalities. Therefore, the Kansas Department of Transportation (KDOT) has implemented several different types of countermeasures to reduce the number of motor vehicle fatalities associated with lane-departure crashes. This research was conducted to estimate the safety effectiveness of commonly used lane-departure countermeasures in Kansas on all crashes and lane-departure crashes using Crash Modification Factors (CMFs). Paved shoulders, rumble strips, safety edge treatments and median cable barriers were identified as the commonly used lane-departure countermeasures on both tangent and curved road segments while chevrons and post-mounted delineators were identified as the most commonly used lane-departure countermeasures on curved road segments. This research proposes a state-of-art method of estimating CMFs using cross-sectional data for chevrons and post-mounted delineators. Furthermore, another state-of-art method is proposed in this research to estimate CMFs for safety edge treatments using before-and-after data. Considering the difficulties of finding the exact date of implementation of each countermeasure, both cross-sectional and before-and-after studies were employed to estimate the CMFs. Cross-sectional and case-control methods, which are the two major methods in cross-sectional studies were employed to estimate CMFs for paved shoulders, rumble strips, and median cable barriers. The conventional cross-sectional and case-control methods were modified when estimating CMFs for chevrons and post-mounted delineators by incorporating environmental and human behaviors in addition to geometric and traffic-related explanatory variables. The proposed method is novel and has not been used in the previous cross-sectional models available in the literature. Generalized linear regression models assuming negative binomial error structure were used to develop models for cross-sectional method to estimate CMFs while logistic regression models were used to estimate CMFs using case-control method. Results showed that incorporating environmental and human-related variables into cross-sectional models provide better model fit than in conventional cross-sectional models. To validate the developed models for cross-sectional method, mean of the residuals and the Root Mean Square Error (RMSE) were used. For the case-control method, Receiver Operational Characteristic (ROC) was used to evaluate the predictive power of models for a binary outcome using classification tables. However, it was seen that the case-control method is not suitable for estimating CMFs for all crashes since the range of the crash frequency is wide in each road segment. A regression-based method of estimating CMFs using before-and-after data was proposed to estimate CMFs for safety edge treatments. This method allows researchers to identify the safety effectiveness of an individual CMFs on road segments where multiple treatments have been applied at the same time. Since this method uses road geometric and traffic-related characteristics in addition to countermeasure information as the explanatory variables, the model itself would be the Safety Performance Function (SPF). Therefore, developing new SPF is not necessary. Finally, the CMFs were estimated using before-and-after Empirical Bayes method to validate the results from the regression-based method. The results of this study can be used as a decision-making tool when implementing lane-departure countermeasures on similar roadways in Kansas. Even though there are readily available CMFs from the national level studies, having more localized CMFs will be beneficial due to differences in traffic-related and geometric characteristics on different roadways.

This Thesis deals with the methods for detection of the outliers and influential points based on method of least squares. The first part of the thesis summarizes the teoretical findings of the method of least squares and both methods for detection of the outliers and influential points based on the method of least squares and also based on robust regression. The practical part of this thesis deals with the application of classic methods for detection of the outliers and influential points on three types of datasets (artifical data, data from specialized literature and real data). The results of the application are subject to qualitative comparisson with the results produced by the methods for detection of the outliers and influentials point based on the robust regression.

Differential Item Functioning Analyses investigates whether individuals with same ability in different groups also show similar performance on an item. In matching the individuals of the same ability, most of the methodologies use total scores of the tests which are usually constructed to be unidimensional. th purpose of the present study is evaluating the PISA 2003 mathematics literacy items through the use of DIF methodology which uses a multidimensional approach in matching students instead of single total score, improve the matching for DIF analyses. In the study factor structure of the tests will be determeined via both exploratory and confirmatory analyses in a complimentary fashion. then DIF analyses conducted using Logistic regression (LR) and Mantel -Haenszel methods.Analyses showed that the matching criterion improved when multivariate analyses were used. the number of DIF items was decreased when the matching criterion is defined based on multiple criterion scores such as mathematical literacy and problem solving scores or two different mathematical literacy subtest score. In addition, qualitative reviews and examination of the distribution of DIF items by content categories, cognitive demands, item types,item text, visual-spatial factors and linguistic properties of items were analyzed to explain the differential performance. Curriculum, cultural and translation differences were the main criteria for the qualitative analyses of DIF items. The results imply that curriculum and translation differences in items might be causing the DIF across Turkish and English versions of the tests.

本研究嚐試找出果凍的最適配方，利用品質機能展開將顧客需求和技術項目結合，找出顧客所重視的技術項目，再依此規劃實驗收集資料。以田口方法和迴歸分析法對數據分析，以提出最適配方。 利用田口方法和迴歸分析法兩種方法找尋最適配方，發現依照不同品質特性所找出的最適配方結果是不相同的，而同時考慮兩種品質特性整體損失函數所找出的最適配方，才能使整體期望損失估計值最小（小於兩個品質特性各別所找出最適配方期望損失估計值的加總）。 由於產品時常有多個品質特性，建議找最適配方時，能同時考量各品質特性的總損失，唯有如此才能找到真正使損失最低的最適生產組合。 / The research goal is to try to find optimal setting combination of one product. We take advantage of QFD (Quality Function Deployment) method to combine the needs of customers and producing technologies; by this, we can tell which technologies are more important for customers’needs. After knowing which technologies are important, we could order an experiment to collect raw data. Using Taguchi Method and Regression Analysis Method to analyze the raw data, we can extract optimal solutions. We found that we could get different optimal solution by different character of a product. Therefore, only considering all characters of a product at the same time, we can get the really optimal setting combination; or we just could get sub optimal setting combinations.

Ce travail porte sur les séries spatiales. On étudie les phénomènes dont l’observation est un processus aléatoire indexé par un ensemble spatial. Dans cette thèse on s’intéresse aux données bidimensionnelles régulièrement dispersées dans l’espace, on travaille alors dans un rectangle régulier (sur Z2) . Cette modélisation vise donc à construire des représentations des systèmes suivant leurs dimensions spatiales et à ses applications dans de nombreux domaines tels que la météorologie, l’océanographie, l’agronomie, la géologie, l’épidémiologie, ou encore l’économétrie etc. La modélisation spatiale permet d’aborder la question importante de la prédiction de la valeur d’un champ aléatoire en un endroit donné d’une région. On suppose que la valeur à prédire dépend des observations dans les régions voisines. Ceci montre la nécessité de tenir compte, en plus de leurs caractéristiques statistiques, des relations de dépendance spatiale entre localisations voisines, pour rendre compte de l’ensemble des structures inhérentes aux données. Dans la plupart des champs d’applications, on est souvent confronté du fait que l’une des sources majeures de fluctuations est la saisonnalité. Dans nos travaux on s’intéresse particulièrement à ce phénomène de saisonnalité dans les données spatiales. Faire une modélisation mathématique en tenant en compte l’interaction spatiale des différents points ou localités d’une zone entière serait un apport considérable. En effet un traitement statistique qui prendrait en compte cet aspect et l’intègre de façon adéquat peut corriger une perte d’information, des erreurs de prédictions, des estimations non convergentes et non efficaces. / This thesis focuses on the time series in addition to being observed over time, also have a spatial component. By definition, a spatiotemporal phenomenon is a phenomenon which involves a change in space and time. The spatiotemporal model-ling therefore aims to construct representations of systems taking into account their spatial and temporal dimensions. It has applications in many fields such as meteorology, oceanography, agronomy, geology, epidemiology, image processing or econometrics etc. It allows them to address the important issue of predicting the value of a random field at a given location in a region. Assume that the value depends predict observations in neighbouring regions. This shows the need to consider, in addition to their statistical characteristics, relations of spatial dependence between neighbouring locations, to account for all the inherent data structures. In the exploration of spatiotemporal data, refinement of time series models is to explicitly incorporate the systematic dependencies between observations for a given region, as well as dependencies of a region with neighboring regions. In this context, the class of spatial models called spatiotemporal auto-regressive models (Space-Time Autoregressive models) or STAR was introduced in the early 1970s. It will then be generalized as GSTAR model (Generalized Space-Time Autoregressive models). In most fields of applications, one is often confronted by the fact that one of the major sources of fluctuations is seasonality. In our work we are particularly interested in the phenomenon of seasonality in spatiotemporal data. We develop a new class of models and investigates the properties and estimation methods. Make a mathematical model taking into account the spatial inter-action of different points or locations of an entire area would be a significant contribution. Indeed, a statistical treatment that takes into account this aspect and integrates appropriate way can correct a loss of information, errors in predictions, non-convergent and inefficient estimates.